Cook, John A. Time and The Biblical Hebrew Verb The Expression of Tense, Aspect, and Modality in Biblical Hebrew
Cook, John A. Time and The Biblical Hebrew Verb The Expression of Tense, Aspect, and Modality in Biblical Hebrew
Cook, John A. Time and The Biblical Hebrew Verb The Expression of Tense, Aspect, and Modality in Biblical Hebrew
J ohn A. C ook
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Abbreviations and Symbols . . . . . . . . . . . . . . . . . . . . . . . . xiii
Chapter 1. A Theory of Tense, Aspect, and Modality . . . . . . . . . . 1
1.1. Pre-modern Treatments of Tense and Aspect 1
1.2. The R-Point and Modern Tense Theory 4
1.2.1. Jespersens Universal Tense Theory 4
1.2.2. Reichenbachs R-Point Theory 7
1.2.3. Revisions of Reichenbachs R-Point Theory 10
1.2.3.1. Hornstein: Reducing R-Point Redundancies 10
1.2.3.2. Bull: Multiple R-Points 12
1.2.3.3. Comrie: A Compromise R-Point Theory 13
1.2.3.4. Declerck: A Bifurcated R-Point 15
1.2.4. Summary: The R-Point and Tense Theory 18
1.3. A Primer on Aspect 18
1.3.1. Situation Aspect 19
1.3.2. Phasal Aspect 25
1.3.3. Viewpoint Aspect 26
1.4. The R-Point in Tense-Aspect Theory 28
1.4.1. M. R. Johnson: The Triadic Relationship of E, R, and S 28
1.4.2. W. Klein: The R-Point as Topic Time 31
1.4.3. Olsen: Multiple R-Points 33
1.4.4. Summary: The R-Point and Tense-Aspect Theory 36
1.5. The R-Point and Discourse 36
1.5.1. Explaining R-Point Movement with Viewpoint Aspect 37
1.5.2. Explaining R-Point Movement with Situation Aspect 38
1.5.3. Explaining R-Point Movement with (Un)boundedness 41
1.5.4. Summary: The R-Point in Discourse 41
1.6. A Primer on Modality 42
1.6.1. Backgrounds of Modality: Grammar, Logic, and
Speech Acts 42
1.6.2. Defining Modality 45
1.6.3. Categorizing Modality 47
1.6.3.1. Classifications of Modalities 47
1.6.3.2. Modality and Speech Acts 52
1.6.3.3. Modality and Mood Systems 53
1.6.4. Conclusion: Modality and TAM Systems 54
vi
Contents
1.7. A Model of Tense, Aspect, and Modality 55
1.7.1. Semantics and TAM 55
1.7.2. Time and Space: Temporality and TAM 57
1.7.3. Aspect and Event Time 57
1.7.3.1. Situation Aspect and Event Time 58
1.7.3.2. Phasal Aspect and Event Time 62
1.7.3.3. Viewpoint Aspect and Event Time 65
1.7.4. Tense and Unilinear Time 68
1.7.5. Modality and Alternative Times 70
1.7.6. The Interaction of Tense, Aspect, and Modality 72
1.8.Conclusion 75
Contents
vii
viii
Contents
Preface
A cognate duality marks the coexistence of language and of time. There is a
sense, intuitively compelling, in which language occurs in time. Every speech
act, whether it is an audible utterance or voiced innerly, takes timeitself
a suggestive phrase. It can be measured temporally. . . . But this occurrence of
language in time is only one aspect of the relation, and the easier to grasp. Time,
as we posit and experience it, can be seen as a function of language, as a system
of location and referral whose main co-ordinates are linguistic. Language largely
composes and segments time. (Steiner 1998: 13536)
The origins of this book reach back to my dissertation (Cook 2002), though it
has grown and morphed significantly in the intervening decade. Some portions
of this work bear striking resemblance to its origins, such as the discussion of
the history of tense theory in chap.1 and the discussion of Biblical Hebrew
verb theory in chap. 2, which has mainly been expanded to treat recent research. By contrast, the discussion in chap.3 has been completely revised and
rearranged, at times leading to the rejection of my earlier claims; and the latter
portion of chap.4 represents a framework not even conceived of at the beginning of my interest in the Biblical Hebrew verbal system.
The Biblical Hebrew verbal system continues to exercise scholars, and
though I have attempted to give an accurate picture of the contours of the
discussion, there are undoubtedly various works that I have missed mentioning directly. I trust, however, that I have interacted enough with the range of
approaches to the perennial questions on the Hebrew verb to have done them
justice. With respect to the perennial questions, my answers may appear deceptively traditional. For instance, my answer to the long-standing debate of
tense or aspect is not drastically different from Ewalds, though our understanding of the perfective : imperfective opposition encoded in Biblical Hebrew qatal and yiqtol is distinguished by the century and a half of intervening
discussion and the rise of modern linguistics. One distinguishing sign is my
employment of the phrase aspect prominent ( la Bhat 1999) to describe the
Biblical Hebrew verbal system. The framing of the central question as tense
or aspect (or modality) wrongly implies a mutual exclusivity among these
domains of meaning. At the same time, it is unhelpful to claim that they can
never legitimately be separated out in the course of analysis: dissection of the
system does not necessarily imply that any part of the system can function
independently.
ix
Preface
The aspect-prominent verbal system of Biblical Hebrew consists of the following forms treated here: a Perfective (qatal) : Imperfective (yiqtol) opposition; a Past Narrative verb (wayyiqtol); a progressive construction consisting
of the predicative Participle; irrealis mood expressed by the two central forms
exhibiting verb-subject word order, the Irrealis Perfective/qatal (i.e., wqatal)
and Irrealis Imperfective/yiqtol; and a Directive-Jussive subsystem of irrealis
mood, which aligns with the irrealis verb-subject word order. As with almost
any of the worlds verbal systems, this aspect-prominent system can express a
wide range of aspectual, tensed, and modal meanings. The argument of chap.3
is that all these forms can be semantically identified as listed above and that
their expressions of specific aspectual, tensed, and modal meanings are explicable with reference to their general meanings.
Methodologically, I eschew statistical means of validation, pointing out their
weaknesses along the way, and draw on diachronic typology and grammaticalization as an external means of validating my theory. Linguistic typology
consists of collecting data from multiple, genetically unrelated languages and
making generalizations about linguistic structures across those languages. Diachronic typology provides a historical dimension by generalizing not simply
about language structures but about the types of structural changes that are evident in languages over time. Grammaticalization is an offshoot of diachronic
typology that is specifically interested in the structural shift of lexical items to
grammatical items or shifts of grammatical items to become more grammatical
over time. These fields have provided a wealth of data on verbal systems and
diachronic changes to these systems in the worlds languages, which serve as
an external means of validating my theory of the Biblical Hebrew verbal system: for my theory to be valid, it should accord generally with what is known
about verbal systems and the ways that they change over time. Given the inescapable diachronic dimension that is part of studying the ancient, composite
corpus of the Hebrew Bible, the findings of diachronic typology with regard
to verbal systems is a particularly powerful means of escaping the subjectivity
and translation-based approaches of other theories.
Because diachronic changes leave traces on synchronic linguistic structures,
some linguists (particularly those working with grammaticalization theory)
have argued that the diachronic-synchronic distinction is unhelpful and that a
panchronic approach to the linguistic system must be pursued. In this regard,
my work here represents a step back from my embracing of grammaticalization
theory and panchrony in my earlier work (Cook 2001; 2002). In the intervening decade, I have come to recognize that my theory is neither strictly nor consistently in line with grammaticalization theory and the notion of panchrony.
During this time, I have also become convinced that the idea of panchrony
is theoretically problematic (Newmeyer 1998: 290) and that the results of an
approach of this sort are unhelpful (Andrason 2010; 2011a): the inevitably
Preface
xi
Acknowledgments
I have benefited greatly from colleagues working in cognate areas (e.g.,
Semitic languages, Rabbinic Hebrew, linguistics, etc.) and on the Hebrew
verb specifically, both in published form and in various other interactions with
them. I have especially valued feedback from numerous conference papers
over the past decade, by means of which I was able to refine my thinking on
these matters. I trust that I have adequately and fairly portrayed the viewpoints
of my colleagues in these pages.
Several individuals have been especially instrumental in my thinking about
these topics and thus deserve specific mention. Cynthia L. Miller-Naud
guided my initial thinking well on these matters as a dissertation mentor. I was
fortunate to receive the input of Michael P. OConnor before his tragic death,
both when I submitted my proposal to the LSAWS series and in subsequent
conversations with him. Robert D. Holmstedt has been a constant conversation
partner in Biblical Hebrew linguistics since our graduate days, and there is
little of this book that has not already been the topic of our informal interactions and collaborative enterprises. Also, I would be remiss not to mention the
specific help that Joe Salmons of the University of Wisconsin provided me,
first through his service on my dissertation committee and subsequently by
helping me decipher some particularly difficult eighteenth- and nineteenthcentury German passages. Similarly, James Spinti graciously double-checked
my translations of Latin passages. Finally, I want to thank Beverly McCoy,
whose keen editorial sense has contributed significantly to the readability of
this book; and to Jim Eisenbraun and the rest of the Eisenbrauns staff go my
thanks for your always-exemplary work producing good books of mine and
others manuscripts.
My deepest gratitude is reserved for my family: for my soul-mate Kathy,
, without whose support in innumerable ways I could
never have begun, much less finished this work; and for our sons, Jared, Colin,
Tage, and Evan () , who have selflessly allowed
this project to accompany me during the majority of their childhood years.
xii
first person
second person
third person
acc
accomplishment
ach
achievement
act
activity
adj. adjective
asv
Authorized Standard Version (1901)
b.c.e.
before the Common Era
BH
Biblical Hebrew
BHVS
Biblical Hebrew verbal system
coh
cohortative
cond
conditional (protasis)
CS
Central Semitic
decl
declarative (neutral epistemic)
dir directive
DRT
Discourse Representation Theory
dyn dynamic
EA El-Amarna
epis epistemic
ES
East Semitic
ET
English translation
exist
existential predicate
f feminine
fin
final (apodosis/purpose/result)
Fr. French
fut
future time/tense
Gk. Greek
hab habitual
impv
imperative
ipfv
imperfective aspect
juss
jussive
kjv
King James Version (1611)
Lat. Latin
m masculine
nab
New American Bible (1970)
xiii
xiv
niv
Symbols
,
?
@
*
&
+
++...
+V
=
=
xv
xvi
RP
RTO
RTR
S
SS
SV
t
VS
w
X-VS
Chapter 1
A Theory of Tense,
Aspect, and Modality
This chapter introduces the categories of tense, aspect, and modality
(TAM), which are most frequently associated with verbal forms in language.
The discussion is divided into two parts: the first part is a somewhat historically oriented discussion of these categories, focusing particularly on the role
of the R-point in tense and tense-aspect theories (1.16); in the second
part of this chapter, I present a theory of TAM based on a central temporalspatial metaphor and illustrated with fragments of English grammar (1.7).
This chapter provides a general linguistic background for the discussion of the
Biblical Hebrew verbal system in chap. 2, and a theoretical-linguistic basis for
the analysis of the Biblical Hebrew verbal system in chaps. 34.
1.1. Pre-modern Treatments of Tense and Aspect
Tense is closely associated with time, a view reflected in the frequent use
of a single term for both in some languages (e.g., Gk. , Lat. tempus,
Fr.temps).1 However, tense is not identical with time, thus raising the question of how precisely the two are related. Time is widely conceived of in terms
of the triad past, present, and future, yet verbal systems are rarely, if ever, so
simply structured.2 The widespread and widely varying mismatches between
the number of verb forms in language and these three categories of time have
long exercised grammarians and linguists.
One of the earliest extant grammatical treatises on Greek is the
(Tchn Grammatik), ascribed to Dionysius Thrax (around
1. Aristotle (Int. 3, 16b56, trans. Edghill) states that [a] verb () is that which, in
addition to its proper meaning, carries with it the notion of time/tense. Robins (1997: 33)
cautions that the meanings verb and predicate for may not yet have been clearly
distinguished in Aristotles day.
2. Binnick (1991: 34) remarks that [t]his notion of three times was already old in the
time of the ancient Greeks. . . . We find references to the Three Times in all periods and
nearly all the various branches of the Indo-European language family. The unique status of
the present in this tripartite division was discussed as early as Aristotle (Phys. 6.3, 233b33
234a4, trans. Hardie and Gaye), who took the view that the present does not really take up
any time but is an indivisible point that acts as a boundary between the past and future.
Chapter 1
100b.c.e.).3 In the Tchn, the Greek verb forms are reconciled with the three
times by categorizing the Imperfect, Perfect, Pluperfect, and Aorist forms as
variants (Gk. ) of the past.4 Although the Greek tenses are merely
listed in the Tchn, without commentary, the names given to the variants of
past tense are revealing, as shown in [1.1].5
[1.1] Imperfect= extended
Perfect= lying near
Pluperfect= more than perfect/complete
Aorist= indefinite, or undefined
The influence of the treatment of the Greek verb in the Tchn on later grammatical theory is mediated by Priscian Caesariensis (fifth to sixth century c.e.).
Priscian clarifies and extends what is implied in the Tchns label for the Perfect, (lying near), by defining three of the Tchns four variant
past forms in terms of their temporal distance from the present: the Imperfect
denotes events that have begun but have not yet been completed (thus may
extend into the present), the Perfect denotes events that occurred recently, and
the Pluperfect denotes events that occurred a long time ago (Binnick 1991:
11, 466 n.50). Priscians interpretation of the Tchn became the model for
subsequent European relative-tense theories, the hallmark of which was their
distinction of verb forms in terms of their relative temporal distance.
Stoic grammarians (from ca. 300 b.c.e.) proposed a different resolution of
tense and time in Greek that incorporated an early concept of aspect defined
in terms of (in)completion (see Robins 1997: 36).6 The Stoic schema paired
the Greek verb forms based on the morphological similarity of their stems
3.The Tchn is ascribed to Dionysius Thrax, although much discussion has been devoted over the years to questions about the authenticity of this ascription for at least portions
of the work (see Robins 1995; Swiggers and Wouters 1998: xvxxxi).
4. The Present and Future verb forms were naturally associated with present and future
time, respectively, while the Future Perfect was left out of the catalog, perhaps because it
was rare and regarded by Greek grammarians as peculiar to the Attic dialect (so Robins
1997: 36).
5. The passage in question ( ), along with Krschners German translation, reads (1996: 19899):
, , , .
, , , , . Es gibt drei
Tempora (Zeiten): Prsens (Gegenwart), Prteritum (Vergangenheit), Futur (Zukunft). Von diesen hat das Prteritum vier Sorten: Imperfekt (sich erstreckend), Perfekt (nahe stehend), Plusquamperfekt (mehr als vollendet), Aorist (nicht begrenzt).
6.Thus Robins; but Binnick (1991: 2224) notes that scholars have interpreted the
Stoic idea of (in)completion in various ways, some of which do not involve aspect at all but
relative tense: completed events lie mostly in the past, and incompleted events lie mostly in
the future.
Present/Future
(Present)
I loose/am loosing
(Perfect)
I have loosed
(Future)
I will loose
a. Sigmatic stems are named for the characteristic Greek letter /sigma (Eng. s) in their tense
sufformative.
Present
am (Present)
I love/am loving
amv (Perfect)
I have loved
Future
amb (Future)
I shall love
amver (Future Perfect)
I shall have loved
and defined the tenses in terms of two times (past versus present/future or
non-past) and three aspects (incomplete, complete, and indefinite), as shown
in table1.1.7
The Roman grammarian Marcus Terentius Varro (11627 b.c.e.) adapted
this Stoic schema for Latin, separating the present and future times and eliminating the indeterminate aspect, because Latin has no equivalent to the Greek
Aorist (the Latin Perfect expresses the sense of both the Greek Perfect and the
Aorist). The Varronian schema of the Latin verb is given in table 1.2.
Versions of both the relative-tense and the tense-aspect approaches persisted in Classical and European grammatical traditions up until the twentieth
century, as illustrated by the alternative models of the Latin verb offered by
the contemporary Latin grammars of Madvig (1895) and Allen and Greenough
(1931; 1st ed. 1888). Madvigs relative-tense model of the Latin verb, shown
7. Some scholars have proposed that Dionysius Thrax was influenced by Stoic theories,
because he also organized the Greek conjugations based on their morphological similarities
(Robins 1997: 37). The passage in question (the immediate continuation of the passage in
n.5), along with Krschners German translation (1996: 19899), reads:
, , ,
. Drei von ihnen sind miteinander verwandt: das Prsens
mit dem Imperfekt, das Perfekt mit dem Plusquamperfekt, der Aorist mit dem Futur.
Chapter 1
praeteritum
futurum
scrps
I have written/wrote
scrpseram
I had written
scrpser
I shall have written
scrbam
I will write
scrpturus eram/fu
I was on the point of writing
scrpturus er
I shall be on the point of writing
a. Following tradition, Madvig does not label this first row of tenses, though logically the terms
must be understood as expressing present, past, and future in-the-present (see Binnick 1991: 42).
in table1.3, is the product of a long, slow process of modification of Dionysiuss and Priscians theories (Binnick 1991: 38).
It is almost a fully relative-tense theory in that events are located past, present, and future within all three times (in-the-past, in-the-present, and in-thefuture) without necessarily referencing the present point of speaking, as earlier
relative theories had done (e.g., the Greek Perfect lies near the present). Madvigs model of intersecting times thus contains nine relative-tense categories.
By contrast, Allen and Greenoughs tense-aspect treatment of the Latin verb,
given in table 1.4, follows fairly closely the Varronian tense-aspect schema
given in table 1.2 above (p.3).
1.2. The R-Point and Modern Tense Theory
The decisive break with Priscians relative-tense theory came with Jespersens critique of Madvigs model of the Latin verb (Binnick 1991: 38).
Although Jespersens own model of tense was not widely accepted, it was
an impetus for Reichenbachs relative-tense theory, from which all modern
relative-tense theories derive. Modern relative-tense theories feature in common Reichenbachs innovative reference time (R-point)that is, a temporal
location in addition to that of the event and that of the speakers present (time
of speaking), the ordering of which with respect to the latter two temporal locations defines the various tenses.
1.2.1. Jespersens Universal Tense Theory
Jespersens work is important as a bridge between traditional grammar
and modern linguistics (Binnick 1991: 54). Jespersen offered the first modern critique of relative tense and, although ultimately his alternative model
has proven inadequate, his research advanced tense theory by prompting later
Complete
scrps
I have written/wrote
scrpseram
I had written
scrpser
I shall have written
scholars to address the areas that his own theory did not adequately treat (Binnick 1991: 110).
Jespersen employed Madvigs relative-tense model of the Latin verb as a
foil for his universal tense model (see table 1.3). He offered three criticisms
of Madvigs schema (Jespersen 1924: 25556). First, there are redundancies
in Madvigs model that obscure differences in meaning between various verbal forms and categories. On the one hand, the Future verb form scrbam occurs twice, once as future-in-the-present and again as present-in-the-future,
on the basis of which other categories, such as present-in-the-past and pastin-the-present, might likewise be treated as equivalent. Yet the verb forms in
these categories (i.e., Perfect scrps and Imperfect scrbbam) are clearly not
equivalent. On the other hand, two verb constructions appear in a single categorythe future-in-the-presentthat are not synonymous: scrpturus eram
(with the Imperfect form of the copula verb) and scrpturus fu (with the Perfect form of the copula verb). Second, Jespersen objected to Madvigs tripartite
division of present time on the basis that, like Aristotle (Phys. 6.3, 233b33
234a4), Jespersen treated the present as an indivisible boundary between past
and future times. Third, Jespersen thought that a one-dimensional model of
tense, reflective of a linear conception of time, was preferable to Madvigs
two-dimensional model of intersecting times.
On the basis of these criticisms of Madvigs schema of the Latin verb,
Jespersen (1924: 277) proposed the universal model of tenses schematized in
fig.1.1.8 Because he treated the present as an indivisible point or boundary,
his model includes only seven categories of tense, thus eliminating the redundancy and ambiguity that he drew attention to in Madvigs schema. For all his
criticism of Madvigs relative-tense model, Jespersens before- and after-
categories were derived from relative-tense theory. However, by presenting
8. To be more precise, Jespersens model features universal times, which are associated
with tense forms in a one-to-one fashion (see Binnick 1991: 61).
Chapter 1
A Past
C Future
{
{
TIME
before-past
Aa
past
Ab
after-past
Ac
future
after-future
Cc
post-future
290 n.1). However, there appear to be two other, more crucial reasons why
Jespersen could not fit the perfect tense into his system. First, because Jespersen adhered to the Aristotelian notion of the present tense as an indivisible
point,10 he could not divide the present as he did the past and future, which
would have allowed a before-present time slot, which would be the logical
position for the perfect (Binnick 1991: 62). Second, and relatedly, Jespersen
could not allow such a before-present category in order to maintain the generalization that tense back-shifting in constructions such as indirect speech or
hypothetical statements (e.g., He said, I am rich > He said that he was rich;
I am rich > If I were rich) involve just one step backwards (Binnick 1991:
64). Thus, as Binnick observes (Binnick 1991: 6162), Jespersens absolutized
tense model founders in the very cases that a relative-tense model, such as
Madvigs, handles most elegantly:
It was Jespersens rejection of relative tenses as such that prevented him from
recognizing that the retrospective (perfect or ante-) tensesincluding the present perfectcan be viewed simply as pasts relative to the main divisions of past,
present, and future; and similarly, that the prospective (post-) tensesincluding
the conditionalcan be viewed as futures relative to those same main divisions.
Many uses of such tenses in subordinate structures, which a relative-tense theory
can account for directly, are at best handled indirectly in Jespersens theory.
Chapter 1
}
}
New Name
anterior past
simple past
Traditional
Name
past perfect
simple past
English Example
I had done it
I did it
posterior past
future perfect
simple future
posterior future
simple future
I shall do it
I shall be going to do itb
a. Reichenbach (1947: 298) notes that the English conditional, although it expresses posterior
past, is usually classified as a mood, not as a tense.
b. Reichenbach (1947: 297) suggests the example I shall be going to see him as posterior
future, which speaks not directly of the event E, but of the act of preparation for it. However, out
of context it is unclear whether be going (to) is the main verb or the future auxiliary (see Hopper
and Traugott 2003: 13). The combination of the English future auxiliaries will/shall and be going
to is grammatical, though uncommon (e.g., Please be here right at 3 oclock, because I shall be
going to the airport at that time). Reichenbach (1947: 297) notes that languages such as Latin have
a more direct expression of the posterior future in their first periphrastic conjugation, constructed
of the Future Participle plus the Future of the verb to be: abiturus ero I will be one of those who
will leave. See also Allen and Greenough (1903: 107), amaturus ero I will be about to love.
of intersecting times (i.e., both feature nine tense categories, equally divisable into past, present, and future species) and Jespersens linear model (i.e.,
Reichenbach also conceives of time linearly). What particularly distinguishes
Reichenbachs theory from Jespersens is that it does not locate events on
the time line absolutely with respect to the present time of speaking but always relatively, mediated by the point of reference time R. This is reflected
in Reichenbachs (1947: 297) suggested tense names (second column in table
1.5, above), in which past, present, and future refer to the temporal
placement of R relative to S, and anterior, simple, and posterior refer to
the temporal placement of E relative to R. No direct relationship between E and
S is entertained by Reichenbachs model. Hence, multiple orderings in which
only the temporal relationship of S and E are in question are irrelevant (e.g.,
posterior past and anterior future): Further differences of form result only
when the position of the event relative to the point of speech is considered; this
position, however, is usually irrelevant (Reichenbach 1947: 296).
Despite the redundancy of multiple orderings of S and E and the fact that,
like Madvig, future tense occurs twice in Reichenbachs system, the explanatory power of his theory is impressive. This can be appreciated best by examining its handling of subordinate clauses, the issue on which Jespersens
theory foundered (see quotation from Binnick 1991: 6162 in 1.2.1 above).
Reichenbach (1947: 293) claimed that tense-shifting in subordinate clauses
could be explained with his R-point theory in terms of the principle of the
permanence of the reference point, instead of having recourse to a complicated and often violated sequence of tense rules: We can interpret these [sequence of tense] rules as the principle that, although the events referred to in
the clauses may occupy different time points, the reference point should be
the same for all clausesa principle which, we shall say, demands the permanence of the reference point.11 This principle may be illustrated by the analyses in [1.2] (based on Reichenbach 1947: 293), in which the permanence
of the reference points is indicated by the vertical alignment of R (the time of
speaking, S, also remains permenantly positioned).
[1.2] a.
1[And
when she had tired herself out with trying], 2[the poor
little thing sat down] 3[and cried]. (Carroll 2000: 2425)
first clause:
E1 < R1 < S (anterior past)
second clause:
R2, E2 < S (simple past)
third clause:
R3, E3 < S (simple past)
b. 1[I have not decided 2[which train I shall take]]. (Reichenbach
1947: 293)
first clause:
E1 < S, R1 (anterior present)
second clause:
S, R2 < E2 (posterior present)
c. 1[She felt sure 2[she would catch a bad cold]]. (Carroll 2000: 34)
first clause:
R 1, E 1 < S
second clause: R2 < E2 < S
10
Chapter 1
Furthermore, Prior (1967: 13) pointed out that complex sentences such as I
shall have been going to see the queen appear to require more than one reference point. And if this is the case, observes Prior, then it becomes unnecessary and misleading to make such a sharp distinction between the point or
points of reference and the point of speech; the point of speech is just the first
point of reference. . . . This makes pastness and futurity always relative to some
point of referencemaybe the first one . . . or maybe some other. These criticisms of Reichenbachs R-point theory raise key issues such as how exactly to
define the point of reference, particularly with respect to S and E, and whether
multiple reference points are required to account fully for tense. These issues
as well as the redundancies of Reichenbachs theory have been the focus of
subsequent tense theories.
1.2.3. Revisions of Reichenbachs R-Point Theory
Although numerous tense theories have adopted, adapted, or otherwise included elements of Reichenbachs theory, I focus here on four studies that take
Reichenbachs theory as their starting point and have made significant and/
or influential contributions to relative-tense discussions. My interest is particularly with how they have addressed the three concerns regarding Reichenbachs theory mentioned above (1.2.2): the redundancies of Reichenbachs
orderings of S, E, and R; the question whether multiple reference points are
required to account for certain tenses; and the question how to define the point
of reference (the R-point).
1.2.3.1. Hornstein: Reducing R-Point Redundancies
Hornsteins work (1990) recasts Reichenbachs tense theory within a government and binding framework (see Chomsky 1981). However, what is most
germane to the present study is his solution to the redundancies of S, E, and R
orderings in Reichenbachs list of tenses. Hornsteins approach to the problem
of redundancies begins with demonstrating that there are many more order-
11
E<R<S
E, R < S R, E < S
R < S, E R < E, S R < S < E R < E < S
E < S, R E < R, S
S, R, E S, E, R R, S, E R, E, S E, S, R E, R, S
S, R < E R, S < E
S < E < R S, E < R E < S < R E, S < R
S < R, E S < E, R
S<R<E
12
Chapter 1
13
E(PPV)
has sung
E(PP0V)
sings
E(PP+V)
will sing
E(APV)
will have sung
E(RPV)
had sung
E(RP0V)
sang
E(AP0V) E(AP+V)
zero
zero
E(RP+V)
would sing
E(RAPV)
would have sung
E(RAP0V) E(RAP+V)
zero
zero
Figure 1.2. Bulls model of universal tense categories (adapted from Bull
1960: 31).
14
Chapter 1
Table 1.8. Comries Taxonomy of Tenses with His English Examples
(1994: 455961)
Absolute tense
past
present
future
E<S
E, S
S<E
Relative tense
relative past
relative present
relative future
E<R
E, R
R<E
(In all three above examples, R is contextually [were asked/told] located before S: R < S.)
Absolute-relative tense
pluperfect
future perfect
future in the future
future in the past
(=conditional)
future perfect in the past
(=perfect conditional)
E<R<S
S<E<R
S<R<E
R<E<S
R2 < E < R1 < S Mary left home at eight oclock; she would return an
hour later, by which time her son would have left for
school.
Comrie defined absolute tenses as the tenses that use the present moment
[S] as their reference point [R] (1985: 36; see 1994: 4559). Relative tense
includes a reference point (R) that is contextually determined, often by an absolute tense in an adjacent clause (1985: 56; see 1994: 4560), as illustrated in
table 1.8 above. Finally, absolute-relative tense possesses the combined temporal denotation of an absolute and a relative tense: [A] situation is located in
time relative to some contextually given reference point, while this reference
point is in turn located relative to the present moment, all of this being done by
means of a single tense (1994: 4561, italics added; see 1985: 65).
Reminiscent of Jespersen, Comries taxonomy conspicuously lacks the
present perfect, which he claimed (Comrie 1976: 52) differs not temporally
but aspectually from the simple past: The perfect indicates the continuing
present relevance of a past situation. He explains:
In terms of location in time, however, the perfect is not distinct from the past.
The past tense locates an event in time prior to the present moment. If one were
to provide an analysis of the perfect analogous to that of the pluperfect and
the future perfect, then one would say that the reference point for the perfect
is simultaneous with the present moment, rather than being before the present
moment (as for the pluperfect) or after the present moment (as for the future
perfect). The situation in question would then be located in time prior to this
reference point. In terms of location in time, however, this would give precisely
15
the same result as the past, which also locates a situation as prior to the present
moment. Thus, however perfect differs from past, it is not in terms of time location. (Comrie 1985: 78)
16
Chapter 1
[1.5] [E1Jared arrived at five oclock] [E2after the others had left at four.]
[at five oclock]
[at four]
R2-TR
S (R1-TO)
R3-TR
R2-TO
R3-TO
E2 E1
[leave]
[arrive]
In this example, there are three reference points: R1 is the speech time S, which
serves as the time of orientation (R1-TO) for R2, which it locates or refers to
as before S (R2-TR; also temporally modified by the adverbial phrase at five
oclock); at the same time, R2 serves as the time of orientation(R2-TO), locating
E1 [arrive] simultaneous with it, and R3 before it (R3-TR; like R2, also modified
by an adverbial phrase, at four); finally, R3 serves as the time of orientation
(R3-TO), locating E2 [leave] simultaneous with it (Declerck 1986: 323).
Declercks understanding of the reference point presents another approach to
explaining the compositional nature of tense (cf. Hornsteins theory, 1.2.3.1),
a conception that is implicit even in Reichenbachs R-point theory (1947: 297):
S and E are never directly temporally related to each other; rather, R (as an
RTR) is temporally fixed with respect to S, and R (as an RTO) temporally fixes
E (see table1.7, p.12). At the same time, Declercks theory formalizes what
is grapically implicit in Bulls model (see fig. 1.2, p.13)namely, that the
speech time (S; Bulls PP) is simply the first R (so also Prior 1967: 13, quoted
in 1.2.2. above), which serves as the point of view (i.e., an RTO) from which
another reference time is temporally fixed (i.e., an RTR); this latter reference
point may then serve as a new point of view (i.e., a new RTO; Bulls 0V on a
new axis) from which to fix yet another temporal point.
Consistently fixing E temporally from the vantage point of an R TO removes
logical inconsistencies of an absolute-tense treatment such as Comries, in
which R is eliminated when it is coterminus with S. For example, an absolutetense analysis of Tage was home locates the event of [be home] as before the
speech time (i.e., E < S); however, this analysis is invalidated should Tage
remain at home beyond the time of speech. Alternatively, tense in Tage was
home, analyzed according to Declercks theory, locates the event [be home]
as simultaneous with a temporal viewpoint (i.e., an RTO), which is fixed via an
RTR as prior (past) to the speech time S (the first RTO). Hence, the state [Tage
be home] may continue to hold past the time of speech without any logical
inconsistency, since the event is only indirectly located with respect to the
speech time.
17
E, R1 < R2 << S
E, R << S
R2 < E, R1 << S
R3 < E, R1 < R2 << S
E, R S
E, R, S
S < E, R
S < E, R1 < R2
E, R << S
E, R1 < R2 << S
R2 < E, R1 << S
R3 < E, R1 < R2 << S
E, R S
E, R, S
S < E, R
S < E, R1 < R2
Similarly, before can have two distinct temporal meanings: it may signify
that one time is wholly before another or before and up to another time
(Declerck 1986: 362). According to Declerck, this distinction between two
senses of before corresponds to the distinction between past tense and present perfect, respectively. According to Declerck (1986: 350), contra Comrie
(see 1.2.3.3), the past and present perfect can be differentiated in terms of
tense. Declercks taxonomy of tenses in table 1.9 (translated into Reichenbachs sigla) indicates a threefold distinction for before: for wholly before
(<<), for before and up to (), and before (<), which is ambiguous with respect
to the former two interpretations.
18
Chapter 1
19
Binnick (1991)
Aristotelian aspect
Aktionsart
Aspect
that aspect has been analyzed variously in terms of lexical, grammatical (morphological or syntactic), and semantic distinctions (Binnick 1991: 144). These
inconsistencies and divergences are still evident in recent linguistic works on
aspect, as illustrated in table 1.10. Situation aspect, phasal aspect, and viewpoint aspect are introduced in turn below.
1.3.1. Situation Aspect
Situation aspect describes differences in the temporal structure of events.
The recognition of structural differences among events goes back to Aristotles
observations in Metaphysics:
Of these processes, then, we must call the one set movements [], and the
other actualities []. For every movement is incompletemaking thin,
learning, walking, building; these are movements, and incomplete at that. For it
is not true that at the same time a thing is walking and has walked, or is building
and has built, or is coming to be and has come to be, or is being moved and has
been moved, but what is being moved is different from what has been moved,
and what is moving from what has moved. But it is the same thing that at the
same time has seen and is seeing, or is thinking and has thought. The latter sort
of process, then, I call an actuality and the former a movement. (Metaph.9.6,
1048b2734, trans. Ross)17
20
Chapter 1
Activities
run
walk
swim
push a cart
drive a car
Accomplishments
run a mile
walk to school
paint a picture
grow up
recover from an illness
Achievements
recognize
find
win a race
stop/start/resume
be born/die
21
Distributive
Achievement
State
Activity
b. Partial Ordering (Tree Diagram) (adapted from Mourelatos 1981: 201; cf.
Bach 1986: 6)
situations
states
occurrences
(actions)
processes
events
(activities)
(performances)
developments
punctual occurrences
(accomplishments)
(achievements)
[SQA]
[+SQA]
STATE
PROCESS
EVENT
state of change state of change change of state
[ADD TO]
[+ADD TO]
Activity
+ destroy, create
+
Accomplishment
+ notice, win
Achievement
wink, tap, cough
Semelfactive
a. See also the feature chart based on the discussion of Van Valin and LaPolla (1997: 93) in
Butler (2003: 353). Particularly noteworthy is Diks feature chart (1997: 115) in that it presents
a partial ordering relationship within a feature chart format: [dynamic] distinguishes between
situations and events; [control] (i.e., whether the agent is in control of the situation) distinguishes
between states and position as subcategories of situation, and process and action as subcategories
of event; [telic] distinguishes between dynamism and change as subcategories of process, and
activity and accomplishment as subcategories of action. For a comparative discussion of situation
analyses in functional grammar, see Butler 2003: 337448.
22
Chapter 1
[Stages]
+
+
[Telic]
+
+
2829] remarks).20 The four situation aspects can instead be adequately differentiated on the basis of just two features, as Rothsteins feature chart in
table1.12 demonstrates.21
Rothsteins chart distinguishes accomplishments and achievements from
states and activities with the feature [+telic], just as in Smiths chart (fig.1.3d):
telic events are characterized by an inherent endpoint (Depraetere 1995: 23).
The concept stages has been used to explain dynamicity (see Smith 1997: 19;
Olsen 1997: 35; Verkuyl 1993: 15); however, in Rothsteins chart, the feature [+stages] distinguishes activities and accomplishments from both states
and achievements, in contrast to Smiths chart, in which the feature [+static]
(=[dynamic]) distinguishes states from the other three situation aspects (see
fig.1.3d). In Smiths approach, a third feature, [+durative], is needed to differentiate accomplishments from achievements, which unfortunately leads to
unrealized feature combinations. Rothstein (2004: 12) classifies both states and
achievements as [stages] because they lack a sense of progression, though
for different reasons: states, because they are inherently non-dynamic, and
achievements, because they are near instantaneous.
Linguists have employed various linguistic tests to demonstrate the semantic differences among situation aspects, such as incompatibilities, entailments, and (dis)allowed temporal modifications (see Dowty 1979: 5171;
Binnick 1991: 17378; Verkuyl 1993: 3368; C. S. Smith 1997: 3948; Rothstein 2004: 1428). For example, stative situations do not generally occur in
progressive constructions, as illustrated in example [1.6].
20. The same criticism applies to Olsens feature chart in table1.16 (see p.35 below),
despite the fact that both Smith and Olsen recognize more situation aspects than Vendlers
four (see discussion below this section).
21. Though termed differently, these features are comparable to Verkuyls [SQA] and
[ADD TO] (see fig.1.3c). However, Verkuyls model lacks elegance (similarly to Smiths
and Olsens) in that Verkuyl distinguishes only three situation types with these two features, while up to four types can be distinguished with the two (there is no realization of
[+SQA, ADD TO] in Verkuyls model, which would correspond with Rothsteins definition
of achievement).
[1.6] a.
b.
c.
d.
23
24
Chapter 1
25
semantic and implicature properties of VPs (see below, 1.4.3); a less rigid
approach has been offered by Rothstein (2004: 33), who treats the aspectual
categorization of verbal lexemes as providing constraints on the types of VPs
they may head (see below, 1.7.3.1).
1.3.2. Phasal Aspect
Phasal aspects refer to alterations to one or another of the phases of development through which a situation progresses. Phasal aspect includes specifically defined types such as those given in [1.10].22
[1.10] a. Focus on initial phase:
Inchoative: beginning of a state, e.g., Tage became sick.
Inceptive: beginning of a non-state, e.g., Colin began
writing.
b. Focus on final phase:
Cessative: end of a [telic] event, e.g., Colin stopped
writing.
Completive: completion of a [+telic] event, e.g., Colin
finished writing his report.
c. Alteration of middle phase(s):
Iterative: repetition of a minimal activity event, e.g., Jared
knocked for five minutes.
Habitual: regular pattern of repetition of an event, e.g.,
Evan (always) walks to school.
Continuative: continuation of an event without a pause,
e.g., Jared continued knocking.
Resumptive: resumption of an event after a pause, e.g.,
Colin resumed studying his lesson.
Early treatments of phasal aspect, often termed Aktionsarten or modes
daction, varied widely due to the fact that some linguists approached these aspects in terms of morphology and/or syntax, while other linguists understood
them in semantic terms (so Binnick 1991: 144). These alternative approaches
stemmed from the object languages being studied. Thus, for example, Forsyth
(1970: 2930) was able to relate phasal aspects or procedurals to specific
affixes in Russian that signify various nuances such as iterativity, inceptivity,
durativity, terminativity, and so on (see also Binnick 1991: 145). However, because phasal aspect is nongrammatical, optional, and unsystematic (Binnick
22. Iterative is treated here as the repetition of a minimal activity event rather than the
repetition of a semelfactive, following Rothsteins analysis (2004: 2829), on which, see
above (1.3.1). For other taxonomies, see Binnick 1991: 2027.
26
Chapter 1
27
28
Chapter 1
tense
S
status aspect
E
Figure 1.4. M. R. Johnsons schema of the relationships between E, R, and S
(adapted from 1981: 149).
after she reached the wood she started ramblingtwo events in temporal
succession.
Although the perfective : imperfective opposition is central to the discussion
of viewpoint aspect, the perfect and progressive are also types of viewpoint aspects.26 They interact with tense and the other aspectual types in similar ways
as the perfective : imperfective pair, yet they are semantically distinct from the
latter. The distinction between the perfect and perfective, on the one hand, and
the progressive and imperfective, on the other, is explored below (1.7.3.3).
1.4. The R-Point in Tense-Aspect Theory
The functional similarity of viewpoint aspect and the R-point as the viewpoint from which an event is temporally evaluated has not escaped the attention of linguists working with tense-aspect models of the verb. Declercks
theory, with its ubiquitous R-point, lends itself especially well to a tense-aspect
scheme since tense is determined solely by precedence relationships among
R-points that are associated with an S or an E, thereby making the E-R relationship otiose to the interpretation of tense. The tense-aspect theories discussed
here, while differing in other important respects, all reinterpret the E-R relationship as determining viewpoint aspect, and the R-S relationship as determining tense.
1.4.1. M. R. Johnson: The Triadic Relationship of E, R, and S
M. R. Johnsons 1981 article, based on her earlier dissertation (1977), presents one of the earliest tense-aspect reinterpretations of the R-point, based on
her study of the verb in the Bantu language of Kikuyu. She posited the triadic
relationship among E, R, and S, illustrated in fig. 1.4: the E-R relationship
determines viewpoint aspect; the R-S relationship determines tense; and the
26. Olsen (1997: 16366, 17276; see p.27 n.24) takes a minority position, arguing
that the English Progressive encodes imperfective aspect and the English Perfect, perfective aspect.
final
point
initial
event phase
point
29
development phase
result phase
Using the interval event model in fig.1.5 to define E, Johnson analyzed three
morphologically marked viewpoint aspects in Kikuyu: Completive (=perfective), Imperfect (=imperfective), and Perfect. These viewpoint aspects are
27. Interval semantics was developed by Bennett and Partee (1978) most directly in
response to the imperfective paradox (see 1.3.1) and has become standard in linguistic
discussion. According to interval semantics, the truth value of an event is defined relative to
an interval of time rather than a moment of time (Dowty 1979: 13839); moments of time
are considered primitives by which an interval is measured as consisting of an ordered set of
moments (Cann 1993: 23335; see further the event model in 1.7.3 below).
30
Chapter 1
a.
( )
range of R for
imperfective aspect
range of R for
perfect aspect
b. Completive: R, E
e.g., he built a house
Imperfect: R < E (partially)a
he was building a house
(time prior to completion)
Perfect:
E<R
he had built a house (time
after house was completed)b
a. This is a translation of Johnsons definition (1981: 154): For some t in E, R (<) {t} (the
symbol (<) is equivalent to <); in other words, R precedes some t of E.
b. These examples are taken from the glosses of Johnsons Kikuyu examples (1981: 15455),
which are all in the Kikuyu near past tense.
distinguished from each other in terms of their distinct range of R with respect to one or more phases of E. This is illustrated graphically in fig.1.6:
the Completive has a range of R that is coterminous with the event phase; the
range of R for the Imperfect corresponds to the development phase; and the
range of R for the Perfect corresponds to the result phase.
While Johnsons analyses of the Kikuyu Perfect and Completive aspects
correspond to standard treatments of perfect and perfective, respectively (see
1.7.3.3 below and 1.3.3 above), her analysis of the imperfective is somewhat
unique. Rather than defining the imperfective as excluding both endpoints of
the event, she associates R for imperfective with the development phase, which
extends prior to the initial endpoint of the event (cf. 1.7.3.3 below and 1.3.3
above).
Johnsons treatment of S, R, and E with respect to tense and aspect is typical
of the tense-aspect revisions of the R-point theory. By contrast, her inclusion
of an S-E relationship is singular, motivated by the presence of a morphologically marked distinction in Kikuyu between imminent action and manifest
action (M. R. Johnson 1981: 16162). Unfortunately, Johnson did not explore
the role of the S-E relationship as a universal feature of tense-aspect systems.
However, in a review of W. Kleins tense-aspect model (see 1.4.2 below),
Dahl (1997: 425) argues that the sentences in example [1.12] provide evidence
that the temporal ordering of S and E may determine tense choice in English.
31
Table 1.13. W. Kleins Schema of 1-State Events Illustrated for Past Time
(based on 1994: 1025)a
a. I . . [. . .] . . F
b. [ I . . . . . . . F ]
I . . . . [. . . F]
[I . . .] . . . . F
c. I . . . . . F []
[] I . . . . . . F
R E (and R < S)
R E (and R < S)
E < R (and S R)
R < E (and R < S)
a. The tenses in Kleins examples are past (denoted as a simple precedence relationship between R and S) and perfect (denoted as having an R that includes the time of S within it).
32
Chapter 1
Table 1.14. W. Kleins Schema of 2-State Events Illustrated for Past Time
(based on 1994: 1059)
a. [ ]+++++
b. [ +++]++
c. ++[++]+
[] +++++
R SS of E (R < S)
R SS of E (R < S)
SS of E < R (S R)
R < SS of E (R < S)
topic time contrast, since there is no R (topic time) that exists outside the situation (E); (b)with 1-state situations, there is an external topic time contrast,
since another R may lie outside the situation (E); (c) finally, with 2-state situations, there is an external topic time contrast as well as an internal contrast
between the source state and the target state of the situation (E).
[1.13] a. The book is in Russian (0-state event, no topic time contrast, R is always in E)
b. The book is on the table (1-state event, R contrasts with an
R when the book is not on the table)
c. He put the book on the table (2-state event, R contrasts with
another R, as in b, and the source state [book not on the
table] contrasts with the target state [book on the table] in
the same R)
Klein (1994: 105) claimed that, with respect to 2-state events, English happens to overlap the topic time R with the source state as opposed to the target
state, overlap with which is presumably a possibility. Thus, as shown in table
1.14, R may overlap with the source state (SS) of 2-state events (marked by
; + is target state) in the same way that it overlaps with E in 1-state events,
illustrated above, in table 1.13 (p.31).
Unfortunately, Kleins notion that languages may choose to overlap topic
time R with either the source state or target state in 2-state events is problematic. First, such a distinction could theoretically only apply in the case of the
imperfective (table 1.14a): according to Klein, the topic time R for perfective
includes a portion of both the source state and the target state (table1.14b),
and the perfect and prospective tenses are defined by the scope of topic time R
following or preceding the source state (table 1.14c), respectively.
Second, in the case of an imperfective example such as in table 1.14a, a
topic time overlapping with the target state would be nonsensical. As Dahl observed in his criticism of Kleins model, the more conventional view of 2-state
events is that they consist only of a transition from one state to the other, not
initial (source) and resultant (target) states themselves. Citing the example
33
Past
+
0
0
+
Future
0
0
+
+
Burton left Mecca, Dahl (1997: 420) states that it is hard to see how Burton
may be leaving Mecca after he has crossed the line.
1.4.3. Olsen: Multiple R-Points
Olsen (1997) constructed a universal tense-aspect model illustrated with
respect to the verbal systems of English and Koine (New Testament) Greek.
While Olsen treated the E-R relationship as defining viewpoint aspect (cf.
1.4.11.4.2), she presented a more complex analysis of tense by allowing
the presence of multiple R-points (see 1.2.3). In particular, she employed the
notion of a contextually determined deictic center (C; compare with Declercks RTR, 1.2.3.4 above), which is movable and has S as its default position. The relationship between this deictic center C and R, rather than simply
S and R, determines tense in Olsens theory, as illustrated in table 1.15. Thus,
Olsens theory combines the advantages of treating the R-E relationship as
aspect without losing the power of relative-tense theories to analyze tense in
subordinate structures, such as those in example [1.14] (subscripts refer to
multiple Rs but multiple positions of C): R1 (ate) lies before Cpos1 (at S), but
R2 (had cleaned) lies before Cpos2 (at R1).
[1.14] Before they ate (R1 < Cpos1=S) breakfast, Jared and Colin had
cleaned (R2 < Cpos2=R1) their room.
With regard to the E-R relationship, Olsen (1997: 52, 57 n. 33) used an
event model consisting of a nucleus and a codaterms derived from the phonetic and phonological study of syllables. R intersects E at the nucleus for imperfective aspect and at the coda for perfective aspect, as illustrated in fig.1.7.
The possibility that R should be analyzed as an interval rather than a moment
is not addressed by Olsen.
Olsen (1997: 5152) incorporated situation aspect into this event model
by associating the situational features [+telic] with the event coda, and [+dynamic] and [+durative] with the event nucleus. These three features distinguish
the six different situation types shown in table 1.16 (Olsen 1997: 4850; on
semelfactive and stage-level states, see 1.3.1 above).
34
Chapter 1
Time
Figure 1.7. Mari Olsens formalized notation for viewpoint aspect (adapted
from 1997: 63).
35
For example, Olsen argued that the atelic (activity) property of a statement
such as [1.16a] is demonstrably an implicature meaning because it is cancelable through the addition of a singular direct object, a mile, which makes the
statement [+telic] (accomplishment) statement. By contrast, she argued that a
[+telic] statement is not cancelable, demonstrated by the effect of the addition
of the durational phrase for a week in [1.16b]: the statement is interpreted as
iterativeJared swam a mile multiple times during a week.
[1.16] a. Jared swam (a mile). (unmarked for [+telic])
b. Jared swam a mile (for a week). (marked for [+telic])
The case for privatively marked [+durative] and [+dynamic] is analogous:
these values are not cancelable, whereas events unmarked for these values
may nevertheless be interpreted as durative or dynamic by implicature.
Although Olsens privative treatment of situation types presents some valuable insights to the contributions of verbs and their complements in determining situation aspect, her arguments are not fully convincing. For example, if
the complement a mile in [1.16b] above is made plural (Jared swam miles),
the [+telic] value is canceled, making the VP an activity. Olsen (1997: 33),
however, would presumably explain the effect of the plural miles in the same
way as the addition of for a week in [1.16b] above: the resultant VP is not
properly an activity but an iteration of accomplishments. Although this explanation is plausible in the case of for a week in [1.16b], it is less convincing
for Jared swam miles, given that the verb is transitive, and therefore the noun
phrase miles completes the valency frame of the verb. Similarly, Olsens claim
(1997: 47) that semelfactives may be read as durative is unconvincing: she
paraphrases the example John coughed (once) for 5 seconds as It took John 5
seconds to cough once. At the very least, one would need to judge a construction of this sort as extremely rare and as requiring explicit marking to avoid an
iterative interpretation (see C. S. Smith 1997: 53).
Another weakness of Olsens theory is her treatment of perfective aspect.
First, she confused perfective and perfect, as is evident from her analysis of
36
Chapter 1
37
38
a.
b.
Chapter 1
R S0
|
E1
Cl1
R
S0
|
E1 E2
Cl1 Cl2
Figure 1.8. Transference of the reference point (adapted from Kamp and
Rohrer 1983: 254).
The Pass Simple verb tlphona in [1.18a] transfers the R-point forward to
the subsequent event, as graphically illustrated in fig.1.8 (Cl #=clause). The
natural interpretation is that Maria telephoned directly after Pierre entered.
By contrast, the Imparfait verb tlphonait in [1.18b] does not move the
R-point but presents a situation that temporally spans the time of the Pass
Simple entra, as illustrated in fig.1.9. Thus, Maria is portrayed as being in the
process of telephoning at the time Pierre enters.
Thus, Kamp and Rohrer (1983: 255) concluded that the reference point
about which Reichenbach speaks is established by context and that, rather
than there being multiple reference points, there is just one transient reference
point, the movement of which is effected by viewpoint aspect: a perfective aspect verb transfers the R-point, whereas an imperfective aspect verb does not.
1.5.2. Explaining R-Point Movement with Situation Aspect
In contrast to Kamp and Rohrers viewpoint-aspectual explanation of temporal movement, many of the early semantic studies of discourse explained the
movement of the R-point in terms of situation aspect: achievements and accomplishments move the R-point forward, whereas states and activities do not
(e.g., Dry 1981; Partee 1984; Reinhart 1984; Dowty 1986; Heinrichs 1986).
A more recent scholar to take this tack is ter Meulen, whose dynamic aspect
theory (1995) uses insights from situation semantics.30 The basis of her theory
30. Situation semantics is a theory developed by Barwise and Perry (1983) as an alternative to model-theoretic (Montague) semantics. At the heart of the theory is its analysis of sentences in terms of situationsconsisting of a location, a relation, and a truth valuerather
than truth conditions. Situation semantics takes context into account to a greater degree than
Montague semantics (Akman 2006; Dekker 1994).
39
b.
R
E2 E1
Cl2 Cl1
S0
|
Figure 1.9. Retaining the reference point (adapted from Kamp and Rohrer
1983: 255).
40
Chapter 1
[1.19] a. Pam arrived at the station. The train had left already. She
sat down on the bench.
arrive
& leave
source
sit down
source
source
sit down
attempt
look at
sigh &
be noon
feel ill
s it down
attempt
source
look at
sigh
& be noon
41
42
Chapter 1
43
44
Chapter 1
noncontingent
necessarily true
contingent
possibly true
possibly false
noncontingent
necessarily false
Figure 1.10. Necessity and possibility (adapted from Cann 1993: 27172).
45
46
Chapter 1
t=now
Figure 1.11. Branching timeline (see Hatav 1997: 119; Portner 2009: 233).
47
48
Chapter 1
[1.24] a. Epistemic
Speculative: But it may rain outside [a possible conclusion].
(Carroll 2000: 168)
Deductive: There must be a thunderstorm coming on [the
only possible conclusion based on the available evidence].
(Carroll 2000: 171)
Assumptive: All the ashes will get into it [a reasonable conclusion based on what is known]. (Carroll 2000: 135)
b. Deontic
Permissive: You may try it if you like. (Carroll 2000: 209)
Obligative: And here I must leave you. (Carroll 2000: 213)
Commissive: I shall never, never forget. (Carroll 2000:
135)
c. Dynamic
Ability: I can explain all the poems that were ever invented.
(Carroll 2000: 189)
Volitive: You will observe the rules of battle, of course?
(Carroll 2000: 206)
Simplistically characterizing these modal domains: epistemic consists of
evaluative expressions (possibility, necessity, probability, etc.), deontic encompasses modalities involving obligation, and dynamic includes modalities
of ability and volition. Portner (2009: 13440) departs from this traditional
taxonomy in some significant ways. First, he moves away form the notion of
obligation to priority, which encompasses deontic, bouletic (wish), and
teleological varieties of modality. He (Portner 2009: 135) explains: [T]he
idea behind the term priority is that such things as rules, desires, and goals
all serve to identify some possibility as better than, or as having higher priority than, others. Second, Portner (2009: 136) adds generic expressions to the
dynamic category, noting that expressions such as Spiders can be dangerous
and Spiders will be dangerous represent modal expressions that quantify over
individualsthe generic spider in this case.
For Palmer (2001), the crucial distinction within the taxonomy in [1.24]
is between propositional and event modalities: epistemic modality deals with
assessment of a proposition, whereas deontic (or priority) and dynamic assess
events. The distinction can be illustrated by paraphrasing several of the epistemic and deontic examples from [1.24] as in [1.25] (the underlined portions
represent the propositions [ab] and events [cd] on which the modalities
operate).
49
[1.25] a. But it may rain outside. (i.e., But it may be the case that it
will rain outside)
b. There must be a thunderstorm coming on. (i.e., It must be
the case that a thunderstorm is coming on)
c. You may try it if you like. (i.e., You are allowed to try it)
d. And here I must leave you. (i.e., I am obliged to leave you
here)
Another distinction that cross-cuts the traditional categories of epistemic,
deontic (or priority), and dynamic is the subjective-objective distinction: subjective indicates that the modal force or assessment derives from the speaker,
whereas objective modality is based on some speaker-external source. However, these labels are not consistently applied with these basic meanings.
Rather, at one end of the spectrum, subjective may indicate the category of
modality as a whole in the cases in which modality is defined in terms of the
speakers assessment of propositions or situations (e.g., Palmer 1986: 2). On
the other end of the spectrum, Portner (2009: 137) prefers a limited application
of the terms to the subjective-objective distinction within epistemic modality,
as introduced by Lyons (1977: 79799). Between these extremes stand the discussions regarding the way that the subjective-objective distinction cross-cuts
all three tradition categories of epistemic, deontic, and dynamic (Verstraete
2001; Butler 2003: 46379).38 For example, Verstraete (2001: 1525), based on
his survey of previous theories as well as an analysis of grammatical criteria,
concluded that epistemic is always subjective, that dynamic is always objective, and that deontic may be either subjective or objective.
Lyons (1977: 797) introduced the idea of subjective versus objective epistemic modality with the contrastive pair of interpretations in [1.26bc] for the
sentence in [1.26a] (see Portner 2009: 12223). In the subjective interpretation,
the modal force derives from the speaker him/herself regarding the proposition
that Alfred is unmarried; in the objective interpretation, the modal possibility
is included within the proposition stated by the speaker and thus outside the
speakers assessment in contrast to the subjective interpretation.
[1.26] a. Alfred may be married.
b. Subjective: poss+it-is-so+[Alfred is unmarried]
c. Objective: I-say-so+ poss+[Alfred is unmarried]
38. Functional linguists particularly have given attention to this distinction in the context
of their analysis of utterances as multilevel. Subjective and objective modality are understood as operating on different levels of the utterance (similar to the distinction between
proposition and event modality; e.g., Dik 1997: 24143).
50
Chapter 1
51
from Nuytss (2001: 39) conception of performativity as reflecting the speakers commitment to the proposition. Portners understanding of performativity
might be usefully combined with Verstraetes delineation of deontic modality
(2007) by proposing that performativity is related to the distinction between
objective deontic expressions with a specified source of obligation and those
without: the former are performative, whereas the latter are not (see examples
[1.27bc]).
An alternative sort of taxonomy, as illustrated in [1.29], is represented by
Bybee, Perkins, and Pagliuca (1994: 17680): the subjective-objective distinction discussed above intersects with traditional grammatical moods associated
with deontic modality, manifest as speaker-oriented versus agent-oriented
modalities (see also Bybee and Fleischman 1995: 6).
[1.29] Taxonomy of modalities related to mood-based categories
(based on Bybee, Perkins, and Pagliuca 1994: 17680)
a. Epistemic modality (applies to assertions and indicates the
extent to which the speaker is committed to the truth of the
proposition [1994: 179])
Possibility: But it may rain outside. (Carroll 2000: 168)
Probability: I shouldnt know you again if we did meet.
(Carroll 2000: 194)
Necessity: There must be a thunderstorm coming on.
(Carroll 2000: 171)
b. Speaker-oriented modality (allows the speaker to impose
conditions on the addressee [Bybee, Perkins, and Pagliuca
1994: 179])
Imperative: Go and order the drums to begin. (Carroll
2000: 200)
Prohibitive: Now dont interrupt me! (Carroll 2000: 128)
Hortative: Kitty, dear, lets pretend. (Carroll 2000: 130)
Admonitive: Beware the Jabberwock, my son! (Carroll
2000: 137)
Permissive: You may rest a little now. (Carroll 2000: 146)
c. Agent-oriented modality (reports the existence of internal
and external conditions on an agent with respect to the
completion of the action expressed in the main predicate
[Bybee, Perkins, and Pagliuca 1994: 177])
Obligation (external, social conditions): We must have a bit
of a fight. (Carroll 2000: 170)
52
Chapter 1
53
associated grammatical moods, as Lyons did. On the other hand, one cannot
fail to observe an almost exclusive connection between particular moods and
illocutionary forces, such as the imperative mood for commands. The asymmetrical relationship between illocutionary force and grammatical structure
along with the inherent focus of speech act theory on the social function of
language leads to the conclusion that illocutionary force belongs more to the
domain of pragmatics than semantics (modality). That illocutionary force is
more properly treated by pragmatics is underscored by the fact that it may
be conveyed by pragmatic or conversational implicature. For example, the
illocutionary (pragmatic) force of a polite command is often conveyed by an
interrogative (semantic) mood or grammatical structure: Would you like to
close the door? (as a polite command to the addressee to enter as opposed to
the grammatical meaning of querying the addressees desires).
1.6.3.3. Modality and Mood Systems
Portner (2009: 25863) distinguishes verbal mood, notional mood, and sentence mood. Verbal mood is of primary interest, given the focus of this study
on verbal systems. Portners notional mood includes phenomena such as infinitives and dependent clauses, while his sentence mood refers to clause types and
has connections with Lyons linking of moods with speech acts (Lyon 1977:
745; see 1.6.3.2). Despite these varieties of mood phenomena, as already
stated (1.6.2; Palmer 2001: 4), mood systems are prototypically binary, exemplified by Indo-European languages featuring an indicative : subjunctive
distinction, or the realis : irrealis marking in Native American languages and
the languages of Papua New Guinea (e.g., Chafe 1995; Romaine 1995). The
latter realis : irrealis is preferable terminology for the typological category
of verbal mood inasmuch as it has the advantage of clearly being a technical
term (so Palmer 2001: 4) and not being subject to confusion with the grammatical categories of a particular language, as indicative and subjunctive are
(see Portner 2009: 258). In addition, although Palmer (2001: 5) has claimed
that strictly, there is no typological difference between indicative/subjunctive and realis/irrealis, he recognizes that the binary characterization is really
not wholly appropriate of indicative : subjunctive systems, since other verbal
moods such as imperative and jussive stand outside the indicative : subjunctive opposition. Palmer (2001: 5) goes on to note a number of differences in
the distribution and syntactic functions of the realis : irrealis and indicative :
subjunctive. For one, the subjunctive, in contrast to the irrealis, is mostly found
in subordinate clauses (hence the name subjunctive). A second difference is
that the irrealis marker often co-occurs with other grammatical markers that
are in complementary distribution with the indicative : subjunctive marking,
such as imperative. A third is that both irrealis and subjunctive markers are
often redundant but for different reasons: the irrealis because it co-occurs with
54
Chapter 1
other grammatical markers, and the subjunctive because its occurrence in subordinate clauses is usually determined by the type of complementizer. Finally,
in contrast to indicative : subjunctive systems, realis : irrealis systems do not
generally co-occur with tense systems; rather, past and present will usually be
marked as realis, and future marked as irrealis.
As a semantically based typological (or metalinguistic) category, the binary
realis : irrealis category is difficult to establish and has been criticized as generally unhelpful (so Bybee 1998). Part of the difficulty lies in the syntactic basis
of the distinction in certain cases, such as a irrealis or subjunctive mood being
determined by the particular complementizer. Nevertheless, the general association of realis as referring to real or actualized events or assertions versus
irrealiss association with events that are not necessarily real or actualized, or
are nonassertions remains widely applicable (Mithun 1999: 173; Palmer 2001:
11112; Portner 2009: 1).39
The lack of cross-linguistic uniformity in particular has raised doubts concerning the validity of the typological category realis : irrealis (so Bybee, Perkins, and Pagliuca 1994: 23640). However, Mithun (1999: 173) has pointed
to the relativity of the binary opposition of realis : irrealis in that, while there
is a unifying semantic distinction between realis and irrealis, the expression
of this semantic distinction in different TAM systems varies. This approach
also seems to obviate Bybees (1998: 269) objection to the category irrealis,
concluding that it is simply too general to be useful. While this criticism
may be valid with regard to its notional use (in which it may apply across the
entire range of modalities; see, e.g., Palmer 2001: 16368), the term is useful
as almost a meta-modality category that recognizes a certain affinity among a
wide range of modalities that may be expressed by a single grammatical mood,
though the alignment of modalities and the binary-mood distinction may vary
to some extent among languages.40
1.6.4. Conclusion: Modality and TAM Systems
The sprawling character of modality, the lack of strong uniformity of its
categories across languages, and the relative recency of its study versus tense
and aspect make it difficult to draw firm conclusions from the preceding overview. Relevant to the theoretical discussion in the following section is that the
categories of TAM may all be analyzable in terms of temporality, thus pro39. Mithun (1999: 173) describes the distinction as follows: the realis portrays situations as actualized, as having occurred or actually occurring, knowable through direct perception. The irrealis portrays a situation as purely within the realm of thought, knowable
only through imagination (see also Mithun 1995: 386).
40. An example is generic expressions, which Portner (2009: 45) classifies as a modal
(i.e., irreal) expression but which lacks uniformity across languages; in Biblical Hebrew
genericity is mainly expressed with realis mood constructions (see Cook 2005; Holmstedt
2005).
55
Although Portner is referring to the typological study of modality across languages, the lack of uniformity across languages makes it advantageous to approach modality in specific languages in the same way, such as looking for
obviously modal expressions (e.g., Biblical Hebrew Imperative and Jussive
verbs) and then seeing what other expressions share common features with
these modalities.
1.7. A Model of Tense, Aspect, and Modality
In the second part of this chapter, a model of TAM illustrated with grammatical fragments from English is presented. There are several aims in constructing this model. First, it provides opportunity to distill some of the important
contributions to understanding TAM that were noted in the survey in the first
part of this chapter. Second, it enables further exploration of the interaction
between the parameters of TAM. While it was valuable to treat these three parameters separately, as in the preceding surveys, they must also be understood
in their dynamic interaction in natural language. Finally, this discussion aims
at providing a modest formalization of TAM, thus providing a useful meta
language for analyzing the object language of BH in subsequent chapters.
1.7.1. Semantics and TAM
Linguistic semantics is the field of linguistics that studies meaning in natural language. The field encompasses intersecting interests and ideas from both
logical semantics and philosophical semantics (or philosophy of language; see
Crystal 2008: 428). The most common approach to semantics continues to be
Montague semantics (see Janssen 2006: 7.24455), named after its originator, Richard Montague (1974). Montague argued that the semantics that was
developed for the purpose of interpreting formal languages (i.e., mathmatical
and logical languagues) is equally suitable for providing precise interpretations
of natural language expressions (see Cann 1993: 2).
There are three important principles underlying Montague semantics that
are especially important to introduce here as a backdrop to the semantic theory of TAM to follow. The first of these is the principle of compositionality,
attributed to the German philosopher Gottlob Frege, which states that the
meaning of an expression is a function of the meaning of its parts (Cann
1993: 3). Thus, a compositional semantic theory will explain the meaning of
56
Chapter 1
57
58
Chapter 1
I1
I2
m1 m2 . . . m n
{
m1 m2 . . . m n
. . . In
m1 m2 . . . m n
time
preparatory phase
[InucleusF]
I1
m1 m2 m3 m4 . . . m n
[coda]
resultant phase
time
Figure 1.13. A model of events (I=initial point, F=final point, I#=interval, m#=moment).
do not occur during every moment of their duration, evaluating the truthconditional value of an event at a particular moment in time is inadequate.
According to interval semantics, events must be evaluated at intervals of time,
which comprise primitive moments of time ordered by a precedence relationship (<) (see Cann 1993: 23435). Interval semantics prevents the evaluation
of examples such as Kathy worked all day in the garden as true at the interval all day from being invalidated by the possibility that Kathy took periodic
breaks through the day (see Dowty 1977: 50). A model of event time based on
interval semantics may be schematized as in fig.1.12.
A model of event time is not identical with a model of events. In order to be
able to model events in terms of their location and progression in event time,
one must have a means of designating phases of events. One way of doing this
is to analyze event structure using terminology from syllable phonology: an
event consists of an onset, a nucleus, and a coda (cf. Olsens use of nucleus
and coda, 1.4.3). The onset refers to the preparatory phase of an event, and the
coda to the resultant phase. The nucleus refers to the event proper, which consists of at least one interval (except in the case of achievements; see 1.7.3.1)
and is bounded by an I(nitial) endpoint and a F(inal) endpoint. This event
model is portrayed in fig.1.13.
1.7.3.1. Situation Aspect and Event Time
Rothsteins feature chart, presented above (table 1.12, p.22) and repeated
here (table 1.18), is the starting point for the following analysis of situation
aspect. As mentioned earlier (1.3.1), the strength of Rothsteins feature chart
is its linguistic elegance: only two features are used to distinguish the four
situation types, thus leaving no feature combinations unrealized.
59
[Stages]
+
+
[Telic]
+
+
In contrast to other feature charts, Rothsteins uses the feature [stages] instead of [dynamicity] or [stativity] (cf. Olsens chart in table 1.16, p.35,
and C. S. Smiths chart in fig.1.3d, p.21; see Smith 1997: 19 on dynamic
events having stages). This approach implies a localist understanding of dynamicity, such as Verkuyl (1993: 15) presents: dynamic (or [+stages]) events
feature progress, defined as the movement of a theme from a source to a goal.42
This progression may be more or less metaphorical. For example, in John bicycled home (from the office), the progress is fairly literal, expressible in terms
of the location of John along the way home: initially, he is still at the office <t0,
office>, then he proceeds past the library <t1, library>, then past the park <t2,
park>, and the last point is at home <tn, home> (see Verkuyl 1993: 218). By
contrast, the progress in an event such as Colin ate three sandwiches is more
metaphorical; we can conceive of the progress in terms of Colin moving along
a path of sandwich eating: <t0, no sandwiches eaten>, <t1, half a sandwich
eaten>, . . . <tn, three sandwiches eaten> (see Verkuyl 1993: 239).
Rothsteins [stages] represent more than simply a terminological difference with other feature charts in that she does not identify [+stages] as a property of achievements (cf. Olsens chart in table 1.16, p. 35; Smiths chart
in fig. 1.3d, p. 21; Verkuyls schema in fig. 1.3c, p. 21). This approach
is crucial to the linguistic elegance of her chart, because otherwise, achievements and accomplishments would be indiscernible. Rothsteins reason for not
identifying achievements as [+stages] is because they are near instantaneous
(Rothstein 2004: 12; see also fig.1.14d, p.63).
The property [telic] is a feature of virtually all discussions of situation aspect: accomplishments and achievements are [+telic] because they have an inherent or intended endpoint (Depraetere 1995: 23).43 The [telic] distinction
is rooted in Aristotles observations on and (see quotation in
1.3.1), and has been discussed in the literature in terms of the imperfective
42. The localist theory of aspect argues, partially on the basis of the pervasiveness of
spatiotemporal metaphors discussed above (p.57 n. 41), that temporal relationships are inherently spatial (see Anderson 1973).
43. The atelic : telic distinction is also expressed as homogeneous : heterogeneous and
cumulativity : quantization in the literature (see Rothstein 2004: 810).
60
Chapter 1
61
62
Chapter 1
Though these tests do not enable us fully to predict the shifts among situation types, they are important because of what they tell us about the semantic
character of the feature [telic]: one does not talk about reaching an inherent
endpoint (the feature of telicity) over a period of time (for time) but within
a particular time (in time).
The temporal characteristics of particular situation types can be illustrated
using the event model introduced above in fig. 1.13 (p. 58). While the situation types are distinguished by the features [stages, telic], which apply to
the nucleus of the event model (with the exception of achievements), these feature distinctions may be reflected in the event structure itself, as illustrated by
fig.1.14ad. Other than the notation of a [stages, telic] nucleus, the event
model for states in fig.1.14a is identical with the basic event model in fig.1.13.
The event model for activities in fig.1.14b has a [+stages, telic] nucleus.
The [telic] feature is reflected in the nucleuss arbitrary final endpoint (F arb):
one does not finish an activity but stops at some arbitrary point.
By contrast, the event model of accomplishments in fig.1.14c has a [+stages,
+telic] nucleus and thus a natural final endpoint (Fnat), which when reached
makes the accomplishment finished.
Finally, the event model for achievements in fig. 1.14d lacks a nucleus
through the combination of features [stages, +telic]. Since there are no stages
or progression leading up to the natural final endpoint, the final endpoint becomes a near instantaneous transition from the preparatory phase to the resultant phase of the event (see Dahl 1997: 420).
1.7.3.2. Phasal Aspect and Event Time
Phasal aspect was defined above as introducing alterations to one or another of the phases of development of a situation (1.3.2). The particulars
of these alterations can now be explicated using the event model developed
above (1.7.3): phasal aspects act upon one of the three phases of an event (the
onset, the nucleus, or the coda) or one of the transitions between those phases
(i.e., onset-nucleus, nucleus-coda, or onset-coda in the case of achievements);
phasal aspects transform a given portion of the event into an activity subevent,
and in the case of phasal aspects that act on the nucleus, they also alter the
progression of the event in some way (see example [1.11], p.27).
Other differences between types of phasal aspect have to do with the event
phases and the type of situations to which they apply. For example, one type
of phasal aspect applies to the onset-nucleus transition (or onset-coda for
achievements), illustrated by example [1.33]: the incohative refers to this alteration of states, whereas the inceptive refers to this alteration of dynamic
[+stages] events.
63
[Inucleus[stages, telic]F]
[coda]
I1
preparatory phase
m1 m2 m3 m4 . . . m n
resultant phase
time
b. Activities (Farb=arbitrary final point), e.g., [walk]
[onset]
[Inucleus[+stages, telic]Farb]
[coda]
I1
preparatory phase
m1 m2 m3 m4 . . . m n
resultant phase
time
c. Accomplishments (Fnat=natural final point), e.g., [build a house]
[onset]
[Inucleus[+stages, +telic]Fnat]
[coda]
I1
preparatory phase
m1 m2 m3 m4 . . . m n
resultant phase
time
d. Achievements, e.g., [be sick]
preparatory
phase
[coda]
[stages, +telic]
[onset]
resultant
phase
time
Figure 1.14. Event models for situation types (I=initial point, F=final
point, I=interval, m=moment).
64
Chapter 1
65
[1.35cd], form a pair: both express the extension of an event nucleus, but
the resumptive presumes a pause of unexpressed duration before the event is
continued ([+pause]).
[1.35]
66
Chapter 1
a. Perfective viewpoint
{ }
b. Imperfective viewpoint
{}
the event frame is included in the reference frame. In the case of the perfective,
the reference frame includes an entire interval of the event frame nucleus. 46 By
contrast, the reference frame of the imperfective viewpoint aspect is included
within an interval of the event frame nucleus; it excludes the initial and final
endpoints of the interval (see Hatav 1993).
It is important that the relationship between the event frame and reference frame defining perfective and imperfective aspect be defined in terms
of a single interval of the event frame in order to avoid the mistaken correlation between perfective : imperfect aspects and the notion of complete(d) :
incomplete(d) events. These mistaken implications may be accounted for, and
so avoided, by drawing on Gestalt theory47 and conceptualizing the situation
spatially as in fig.1.15: each arrow represents an interval of the event frame;
the brackets represent the reference frame for each viewpoint.
Perfective aspect includes an entire interval in its reference frame, as the
brackets in fig.1.15a illustrate; thus, it gives the impression that the event is
complete(d). However, the interval in the reference frame is not necessarily the
only interval for which the situation holds, as shown by the broken-line arrows
outside the brackets. By contrast, imperfective aspect excludes the beginning
and end of an interval from its reference frame, represented by the placement
of the brackets in fig.1.15b. As Gestalt theory predicts, when the brackets are
placed over the arrow as in fig.1.15b, one will mentally construct the endpoints of the arrows, represented by the broken lines, thus perceiving that the
arrow (or event) projects beyond the reference frame.
Distance refers to whether the event frame interval(s) is/are discerned
within the reference frame. If an interval is discerned within the reference
frame, this more easily allows for another event to be portrayed as occurring
within that same interval. The perfective aspect has a distant reference frame,
46. This way of defining the relationship between RF and EF avoids the over-richness
produced by Kleins less precise definition of perfective simply as R E (see table1.13,
p.31, above).
47. Gestalt psychology (as it relates to the discussion here) deals with the processes of
perception. It maintains that images are perceived as whole patterns rather than as distinct
parts and formulates principles to explain the perceptual processes, such as the principles
discussed here. However, it is a widespread misconception that Gestalt psychology is only
concerned with a psychology of perception (see Kanizsa 1979: 5571).
67
while the imperfective has a near reference frame. The effect of this difference
in distance is found in the fact that the imperfective lends itself more naturally
to serve as an event frame for another event taking place within the same
interval of time. This is illustrated by example [1.36], repeated from [1.11].
[1.36] a. She was rambling on [ipfv] in this way when she reached
[pfv] the wood. (Carroll 2000: 156)
b. She rambled on [pfv] in this way when she reached [pfv]
the wood.
The example with the English progressive construction (imperfective) in
[1.36a] reports two overlapping events: Alice reached the woods during one
of the intervals of time that she was rambling. By contrast, example [1.36b]
has a temporally successive default reading: Alice rambled on in this way after
the interval of time when she reached the wood.
The perfective : imperfective opposition may be formalized using symbols
from set theory, as illustrated in example [1.37]: the perfective is defined as
RF EFthat is, the reference frame includes an interval of the event frame;
the imperfective is defined as RF EFthat is, the reference frame is included
in an interval of the event frame. A far focal distance is invariably associated
with the wide scope RF EF of the perfective, while a near focal distance is
likewise associated with the scope RF EF of the imperfective.
[1.37] a. She was rambling (ipfv). RF EF(nucleus)
b. She rambled (pfv). RF EF(nucleus)
The notation (nucleus) in the examples refers to the fact that the perfective
and imperfective aspects have a default focus on the nucleus of an event (as
opposed to the onset or coda; cf. the perfect below); this default focus may of
course be altered by phasal aspects (1.7.3.2).
Perfect (also referred to as anterior) is treated here as a viewpoint aspect,
although it is unique compared with the other viewpoint aspects (Comrie 1976:
52). On the one hand, the perfect allows for tense distinctions (e.g., had run,
has run, will have run); on the other hand, it may combine with progressive
viewpoint aspect (e.g., has been sleeping). The ability of the perfect to combine
with the progressive may be explained by analyzing the reference frame of the
perfect as relating in scope and distance to the coda of the event frame rather
than the nucleus (so M. R. Johnson 1981; W. Klein 1994; see fig. 1.6, p.30,
and table1.13c, p.31). Hence, the perfect focuses on the resultant phase of a
prior event nucleus, as illustrated in example [1.38]: the event [promise] is prior
to the reference frame in which the king has promised. The reference frames of
the nonprogressive perfect and progressive perfect contrast with respect to the
event frame coda in the same way as the perfective and imperfective reference
68
Chapter 1
frames relate to the event frame nucleus (i.e., in terms of scope and distance), as
illustrated by the analyses of examples [1.38ab] (cf. [1.37]).
[1.38] a. The king has promised (perf) me. (Carroll 2000: 184)
RF EF(coda)
b. The kitten had been having (perf-prog) a grand game of
romps. (Carroll 2000: 127)
RF EF(coda)
The strength of this analysis is that it captures the current relevance characteristic of the perfect (Binnick 1991: 264), but at the same time it avoids
the problems created by interpreting the event nucleus as extending into the
reference frame (so Hatav 1993: 220; see W. Klein 1994: 104; and table 1.14,
p.32). The perfect focuses on the resultant (or implied) state of a past event.
The resultant state is sometimes semantically connected to the perfect verb
(e.g., He has died [event] > He is dead [state]); oftentimes, however, the resultant event must be determined by real world knowledge (e.g., We cant
come to your party. The police have arrested my wife [event]; thus, My wife is
indisposed. [state]) (see Moens 1987: 7172).
Progressive is analyzed here as a viewpoint aspect. However, it has been
treated variously as a tense, a viewpoint aspect, a phasal aspect, a discoursepragmatic verb, and a modality (see Binnick 1991: 28190; Dik 1997: 225;
Dowty 1977). Aspectually, the progressive is apparently identical with the imperfective (see p.27 n. 24). Nevertheless, there are cross-linguistic features
that differentiate the two verb forms. First, progressives often are or develop
from periphrastic constructions and/or are based on nominal forms (Bybee,
Perkins, and Pagliuca 1994: 130; Dahl 1985: 91). Second, progressives are
more restricted than imperfectives: they generally do not occur with stative
predicates, and they generally have a narrower future time use, expressing either an expected event (Comrie 1976: 3334; Bybee, Perkins, and Pagliuca
1994: 24950) or an element of intention (Binnick 1991: 289). Third, the perfective : imperfective opposition is often correlated with the tensed past : nonpast opposition, whereas the progressive is freely used for past, present, and
future time reference (Dahl 1985: 9293). Hence, although the imperfective
and progressive are semantically indistinguishable, they can be differentiated
based on consistent cross-linguistic characteristics.
1.7.4. Tense and Unilinear Time
Tense is defined by temporal precedence relationships, thus requiring a
simple unilinear spatial metaphor of time. The defining characteristic of the
tense-aspect R-point theories surveyed above is that they use the notion of a
reference point/time for both tense and viewpoint aspect (1.4). A difficulty
69
encountered with this approach is that R is defined differently for analyzing aspect and for analyzing tense. In particular, M. R. Johnsons theory encountered
difficulties over whether to treat R as an interval of time or a point (moment)
of time or both, depending on whether it is associated with the interval E or
the moment S (1.4.1).
What is required to overcome this difficulty is a model that associates an
interval R, whose relationship with the event time determines viewpoint aspect, with a momentary R, whose precedence relationship with S determines
tense. The solution adopted here draws on Declercks bifurcation of R into
RTR (R time referred to) and RTO (R time of orientation; see 1.2.3.4) and
Olsens concept of a deictic center (C; see 1.4.3): tense is determined by a
precedence relationship between S and C, the latter of which is associated with
a relevant RF (denoted as CRF), whose relationship with EF determines viewpoint aspect. Thus, analysis of the basic tense distinctions may be expressed as
in example [1.39] (taken from table 1.8, p. 14).
[1.39] a. John ate an apple (past). C < S
b. John is eating an apple (pres). S,C
c. John will eat an apple (fut). S < C
However, Dahl has criticized the R-point tense-aspect approach for neglecting the temporal relationship between the event time and the speech time,
which had been the basis for defining tense prior to the R-point theory (see
1.4.1). Dahl (1997: 425) introduced the contrastive examples [1.40ab] to
argue that the event timespeech time relationship can be determinative in
tense choice. An adequate analysis of these examples requires that tense be
defined compositionally () as consisting of a CRF-S precedence relationship
and an EF-S precedence relationship, as illustrated by the analyses accompanying examples [1.40ab].
[1.40] a. Today, my office hours are from ten to twelve.
[CRF(today), S] [S EF(ten to twelve)]
b. Today, my office hours were from ten to twelve.
[CRF(today), S] [EF(ten to twelve) < S]
In these examples, the CRF-S relationship remains constant (defined precisely
by the temporal adverb today), whereas the EF-S relationship contrasts in the
two examples, explaining the variation in tense choice. In order to handle tense
adequately in expressions requiring multiple R-points, the deictic center must
be defined as a transient time that is initially located relative to the speech
time and ultimately associated with a reference frame but meanwhile may be
located by the tense at some other temporal position.
70
Chapter 1
t=now
e1
mic
piste
e
=
R R=deontic
e2
R
R = dynam
ic
=
al
po
stu
e 3 dition
lat
n
ed
co
=
R R=final
e
4
e5
e6
Another difficulty that confronts both R-point tense and tense-aspect theories is the issue of multiple R-points (see 1.2 above). The function of this
transient deictic center in defining tense may be illustrated on the subordinate
conditional constructions in example [1.41]. The conditional would denotes
the event [walk] as occuring after some time determined by the context
namely, the past tense verb said of the matrix clause. Thus, the deictic center
is initially located left by the past tense said (Cpos1); from there, the conditional
would walk locates C rightward to a position between Cpos1 and S, where it is
associated with RF (CRF) (EF is related to RF to determine viewpoint aspect:
RF EF).
[1.41] He said he would come in. (Carroll 2000: 225)
Cpos1 < CRF < S & RF EF
1.7.5. Modality and Alternative Times
In keeping with a unified analysis of TAM in terms of temporal relationships, we defined modality in terms of the temporal existence of an event or
with respect to alternative situations (1.6.2). For the purposes of this book, it
is unnecessarily complicated to introduce a full modal logic model (for which,
see Allwood, Andersson, and Dahl 1977: 10824; McCawley 1993: 372416;
Portner 2009: 946); however, the principles of modal logic underlie the following simplified formalization of modality in terms of sentential operators.
Most basically, the discussion assumes that modality is explicable in terms of
accessibility relations and the necessity and possibility modal operators (see
Portner 2009: 31).48
48. For example, a modal logic expression of the truth conditions for It may snow might
be formulated as follows: A=[snow]; A is true in a world w if for some w, such that
RKww, A is true in w. (RK stands for an epistemic [K=knowledge] accessibility relation
between alternative worlds or situations.)
71
Absolute modalities
Declarative (neutral epistemic)
decl
Epistemic
Obligation
Directive
epis
Volitive/optative/commissive
Dynamic
vol
obl
dir
dyn
Relative modalities
Conditional (protasis)
cond
Final (apodosis/purpose/result)
fin
72
Chapter 1
view that decl is always subjective and dyn never (see 1.6.3.1).50 The use of
and vol attempts to capture the distinction between speaker-oriented and
agent-oriented deontic modalities as given in example[1.29] above (p.51).
In contrast to my previous presentation (Cook 2008b: 10), here I have excluded
causal and concessive from the relative (or contingent) modalities because unlike conditional and final clauses the postulated situation is often realis, though
it does not need be, as illustrated by the the following realis and irrealis causal
expressions in examples [1.42ab].
dir
73
The formal analysis of example [1.44] defines the TAM values in turn,
separated by ampersand (&): the tense is marked as past (CRF < S) by the past
auxiliary verb might (see 1.7.4); the viewpoint aspect is perfect (the been
indicating passive and not progressive here), so that the RF includes within its
scope the coda of the activity EF (RF EF(coda)ACT) (1.7.3);51 the modal
verb might (past tense of may) relativizes the truth of the event (e) in terms of
ability (dynamic modality) with respect to some alternative situation (1.7.5).
Linguists have drawn attention to certain constraints on the interaction of
TAM values. Bache (1995) has presented one of the most comprehensive theoretical discussions on the interaction among tense, viewpoint aspect, and situation aspect. He worked with taxonomies of three tenses (past, present, future),
two aspects (perfective and imperfective), and an idiosyncratic partial ordering
of situation types (Bache 1995: 194204, 31319). In addition, he allowed that
expressions may lack any of these paramters (i.e., tense, viewpoint aspect,
situation aspect). On the one hand, Baches discussion draws attention to the
semantic irregularlities that arise from certain combinations, which is helpful
to the extent that it furthers our understanding of these particular TAM values.
On the other hand, Bache has made claims regarding the incompatibility of
certain TAM values that need to be assessed critically.
To begin with, Bache (1995: 28586) claimed that minimal activity events
(or semelfactives) default for imperfective aspect, resulting in an iterative
event; C. S. Smith (1997: 53) also pointed out this phenomenon. This combination is notable in that it shows that iterative phasal aspect may often be
derivative, brought about through the combination of a particular situation
(sub)type and viewpoint aspect. However, it is misleading to say that behaviorally minimal activity events default for imperfective aspect, because iterative
activities may also be expressed by perfective verbs (e.g., Jared knocked for
five minutes, example [1.10c], p.25). A more accurate claim would be that
minimal activity events cannot be expressed in combination with imperfective aspect without being interpreted iteratively. This claim stands over and
against Olsens claim (1997: 47), criticized above (1.4.3), that minimal activity events may be interpreted duratively, as in John coughed (once) for 5 seconds, interpreted as It took John 5 seconds to cough once. This phenomenon of
interaction between viewpoint aspect and minimal activity events is related to
the constraint on achievements and imperfective viewpoint: because achievements lack any event intervals (i.e., the [stages, +telic] event type consists
of an instantaneous transition from a preparatory state to the result state), the
imperfective viewpoint focuses on the preparatory interval(s)for example,
He was winning the race focuses on the preparatory period up until the moment of the winning (achievement).
51. A situational aspect label act may be used in place of the feature description [+stages,
telic] of the nucleus or coda (cf. examples [1.301.32], p.61).
74
Chapter 1
Another claim that Bache (1995: 284) made is that states (which he identifies as situation aspect) cannot combine with viewpoint aspectual distinctions; hence, situation aspect (=state) combines with viewpoint aspect. This
also, however, is too strong a claim. While it is true that one linguistic test for
states is that they do not normally appear in the English progressive, this is not
a strict incompatibility, as examples [1.45ab] demonstrate (see Olsen 1997:
3637 for other examples).
[1.45] a. Digory was disliking (stat+ prog) his uncle more every
minute. (Lewis 1955: 20)
b. Yes, we should be knowing their judicial philosophy. We
should be knowing their legal form of reasoning. (Sen.
Charles Schumer, interview by Tim Russert, Meet the
Press, 10 July 2005).
In addition, Baches claim is inapplicable beyond English because incompatibilities or constraints on combinations of states with various viewpoint aspects
appear to be a parameter of language (C. S. Smith 1997: 40).
Bache (1995: 289) along with other linguists (Bhat 1999: 17; Bybee, Perkins, and Pagliuca 1994: 126; C. S. Smith 1997: 185) has claimed that there
is an incompatibility between perfective aspect and present tense. He (Bache
1995: 3058) explains this incompatibility in spatiotemporal terms as being
due to the fact that a certain minimal temporal distance is required to encompass the entire event frame within the reference frame of the perfective
viewpoint; this minimal temporal distance is not afforded by the present tense.
However, it may be more accurate to describe this phenomenon as a constraint than an incompatibility since there is the notable exception of reportative speech, familiar in the setting of sports (especially radio) announcers
(e.g., Jones runs to third base; C. S. Smith 1997: 185). The status of forms
like the Simple Present in English is uncertain (e.g., Jared helps around the
house). These constructions are generally interpreted as habitual or gnomic,
which may be expressed by perfective forms in some languages and imperfective forms in others: in speaking of a repeated total event one can use a
pf. [perfective] verb, thus stressing each individual total event, or use an ipf.
[imperfective] verb, which means that the stativeness of unlimited repetition
takes precedence (Dahl 1985: 79). C. S. Smith (2008: 235) approaches this
phenomenon from the other direction when she proposes a temporal location
default pattern whereby bounded events are temporally located in the past, and
unbounded are located in the present. Perfective aspect, which makes events
bounded (C. S. Smith 2008: 238; see table 4.1, p. 282), therefore defaults
for past temporal interpretation and is constrained from normally being interpreted as present.
75
76
Chapter 1
the event structure (1.7.3); tense was defined in terms of precedence relationships between the speech time and the event time, mediated by a transient deictic center (1.7.4); finally, modality was defined in terms of alternative times
and their relationship with the actual time (1.7.5). Possible constraints on
the interaction of the parameters of TAM were also briefly examined (1.7.6).
The primary aim of this chapter is to provide the requisite background for
the focus of this book, which is the BH verbal system: the first part of this
chapter provides the necessary background for the critical discussion of TAM
in theories of the BH verb in chap.2; the theory of TAM in the second part of
this chapter is the basis for the analysis of the BH verbal system in chaps.34.
Chapter 2
78
Chapter 2
may be enumerated, in particular. The first problem is that the variety of temporal significations expressed by the forms qatal and yiqtol appear to defy a
singular TAM designation for either form.1
The apparently vast range of temporal meanings that qatal and yiqtol can
express is illustrated by the variety of verb forms used to render them in English translations of the Hebrew Bible in examples [2.1ab].
[2.1] a. English verb forms used to translate qatal:
Simple Past: In the beginning God created (
)the
heavens and the earth (Gen 1:1, rsv).
Past Perfect: Now Jacob did not know that Rachel had
stolen ( ) them (Gen 31:32, rsv).
Present Perfect: They have forsaken ( ) the Lord (Isa
1:4, rsv).
Subjunctive: If you had saved them alive, I would not kill
( ) you (Judg 8:19, nrsv).
Present: The lazy person says () , There is a lion outside! (Prov 22:13, nrsv)
Future: They will not hurt or destroy on all my holy mountain, for the earth will be full (
)of the knowledge of
the Lord (Isa 11:9, nrsv).
Future Perfect: He shall die in his iniquity; but you will
have saved ()
your life (Ezek 3:19, nrsv).
Modal: Awake, my God; decree (
)
justice (Psa 7:7[6],
niv).
b. English verb forms used to translate yiqtol:
Simple Past: Then Joshua built (
)an altar in Mount
Ebal to the Lord (Josh 8:30, rsv).
Past Progressive: But a stream was welling up ( )out
of the earth (Gen 2:6, nab).
1. Following a now widespread convention, I avoid the use of traditional, semantically laden labels for the BH verb forms in favor a simplified transcription of the unmarked
paradigm form (3ms of QL to kill, a root traditionally employed in comparative Semitic
studies) to refer to each verbal conjugation (more traditional names are given here in parentheses): qatal (Perfect), yiqtol (Imperfect), wayyiqtol (Waw-consecutive Imperfect), wqatal
(Waw-consecutive Perfect). I have, nevertheless, retained the traditional labels for the Participle, Infinitive, and directive-volitive forms, except for the Cohortative (i.e., Imperative
and Jussive); the accuracy and utility of these latter labels is assessed in chap. 3. Throughout
the discussion, I follow the oft-used linguistic convention of capitalizing the names of BH
and other language-specific verb forms (except for the BH verb forms listed here as referred to by transcription), in distinction from universal semantic categories, which I present
inlowercase.
79
And he asked them, Where do you come from?
b. Josh 9:8
And Joshua asked them, Who are you and where do you
come from?
80
Chapter 2
Table 2.1. Statistics for English Verb Forms Used in the rsv
to Translate qatal and yiqtol (McFall 1982: 18687)
The second problem is that alongside the qatal and yiqtol verb forms are
two other forms, traditionally called waw-conversive or waw-consecutive
forms because of the characteristic presence of the prefixed waw (and) conjunction: wqatal and wayyiqtol. Aside from the phonological peculiarity in
the attachment of the conjunction on wayyiqtol ( / versus the usual shape of
the conjunction ),
these forms are only superficially and partially distinguishable morphologically from their non-waw-prefixed counterparts. The wqatal
form has an ultimate instead of a penultimate accent in some the 2ms and 1s
forms (e.g., /
and you/I will keep versus /
]
]
[and] you/I
kept); but this stress variation is a poor diagnostic because it appears inconsistently on the wqatal form even within a single passage (e.g., Amos 1:45,
78). The wayyiqtol form exhibits an apocopated shape with roots that have
a glide (w/y) in the second or third position (e.g., he built versus
]]
Formal
Similarity
wayyiqtol
(w)yiqtol
81
(w)qatal
wqatal
Semantic Similarity
Figure 2.1. Relationships among qatal, yiqtol, wqatal, and wayyiqtol forms.
is illustrated by verses, such as in example [2.4], in which the two forms are
conjoined.
[2.4] Gen 2:24
Therefore, a man will leave his father and his mother and will
cleave to his wife.
At that time, Berodach-baladan son of Baladan king of
Babylon sent letters and a gift to Hezekiah, for he heard
that Hezekiah was ill.
b. Isa 39:1
At that time, Merodach-baladan son of Baladan king of
Babylon sent letters and a gift to Hezekiah; he heard that he
was ill and recovered.
Schematically, the relationship among these four verb forms is illustrated in
fig.2.1, in which the verb forms are shown as being semantically similar across
the horizontal axis and formally similar along the vertical axis.
The dilemma that this four-way relationship poses for semantically distinguishing among these forms is trenchantly stated by Binnick (1991: 441):
If the waw adds no temporal (tense or aspect) meaning, then the difference between verbs with waw and verbs without waw cannot be a semantic one. But
apparently it is, for the forms with the waw are generally seen as reversing the
values the tenses normally have. To reconcile the two, we must assume that the
82
Chapter 2
forms without the waw and those with it do not in fact differ in semantics, but the
only way this is possible is if the tense forms do not differ from one another in
meaning to begin with.2
Finally, a third problem is that the division between indicative and non
indicative systems in BH is fuzzy. This feature is not unique to BH; many languages have forms that function both with indicative and various nonindicative
meanings. However, in BH this characteristic only exacerbates the problems
outlined above. It is not just an issue of the functional diversity of individual
forms but the apparently unsystematic mixing of forms. For instance, the yiqtol
and Jussive forms in BH are often homonymic; nevertheless, even instances
of yiqtol and Jussive that are morphologically distinct occur with no obvious
difference in meaning, as illustrated by examples [2.6ab].
[2.6] a. Exod 20:10
You shall not do any work/Do not do any work.
b. Gen 22:12
Do not do anything to him.
The wqatal form similarly appears to express both indicative and non
indicative modalities, appearing at times in the same context as indicative
yiqtol (example [2.4] above) but also conjoined to nonindicative Jussive (including Cohortative) or Imperative forms, as in example [2.7].
[2.7] Ruth 2:7
And she said, Let me glean and gather into sheaves behind
the harvesters.
The study of the BHVS is in large part reducible to various attempts to resolve the problems enumerated here. Therefore, the following survey focuses
particularly on proposed solutions to these difficultiesnamely, the semantic
2. Sperber (1943: 199; see also 1966: 59192) took a resigned approach to the dilemma
described by Binnick (above):
Each of these tenses [qatal/wqatal, yiqtol/wayyiqtol] may indicate any and every time.
... The difference that existed between the suffix tense [qatal] and the prefix tense [yiqtol] was not of a temporal, but rather of a dialectic character. . . . The term waw consecutivum, which the grammar had to invent in order to explain the use of an imperfect
[yiqtol] with the meaning of a perfect [qatal], and vice versa, thus becomes obsolete.
For a stringent review of Sperbers grammar, see Blake (1944b) and the response by Edelheit
(1944).
83
distinction between qatal and yiqtol, the relationship between these forms and
their waw-prefixed counterparts, and the indicative : nonindicative distinction.
2.2. The Establishment of the Standard Theory
of the Biblical Hebrew Verbal System
Little justification is required for beginning a survey of theories of the
BHVS with Ewald and S. R. Driver. A little more than a quarter century ago
McFall (1982: 27) observed that the majority of scholars [with respect to their
views of the Hebrew verb] still go back to two 19th-century theories, those of
H. Ewald (1835) and S. R. Driver (1874). Starting with these theories does
not belittle prior contributions but highlights the extraordinary and long-lasting
influence that these theories have had on the discussion of the BHVS.
2.2.1.Before Ewald and S. R. Driver
The impact of the theory of the BHVS established by Ewald and S. R.
Driver is best appreciated against the immediate backdrop of the state of the
field in the early part of the 19th century, just prior to when Ewald introduced
his theory into the field. One of the most pervasive theories at that time was the
waw-conversive theory, which dates back to the Jewish medieval grammarians
(as early as the 10th century, according to McFall 1982: 3) and whose influence is still evident in the labels used for the waw-prefixed verb forms in some
modern grammars.3 The theory is based on an understanding that BH has three
absolute tenses corresponding to the three times, as illustrated in example[2.8]
(see discussion in 1.1):4
[2.8] qatal = Past ()
Participle = Present ( or )
yiqtol = Future ()
On this absolute-tense basis, a conversive explanation was offered for the
two waw-prefixed formswayyiqtol and wqatal. According to the conversive
theory, Hebrew possessed two different waw conjunctionswaw haibbr and
waw hahippk (translated into Latin as waw conjunctivum and waw conversivum, respectively; McFall 1982: 12). The latter type of waw, identified in the
wayyiqtol and wqatal forms, was thought to convert the tense from past to
future or vice versa. One of the clearest expositions of the conversive theory
appears in the grammar of Elias Levita (14681549), first published in 1518.
3. For example, Joon (2006: 117) labels the waw-prefixed forms inverted tenses;
Lambdin (1971: 108) writes of tense values being converted; and even the recent introductory grammars of Kittel, Hoffer, and Wright (2004: 20) and Pratico and Van Pelt (2001:
chap.17) retain the term conversive. See survey in Cook 2008b.
4. For discussion of perspectives on the Participle prior to the 19th century, see Dyk
1994: 36684.
84
Chapter 2
Notice, when you want to convert a past into a future you place a waw with
a ewa in front of it, as in the case of keep in And Yhwh will keep . . . [
:
wqtl.3ms, Deut 7:12], which is like and he will keep [
: yqtl.3ms]. Likewise, And the sons of Israel shall keep the Sabbath [
: wqtl.3mp, Exod
31:16]. It is like and they shall keep [
: yqtl.3mp]. And the waw is always
pointed with ewa except before those consonants that cancel it, turning it into a
ureq, pata, or ireq, just as is explained in the aforementioned passage.
And if you ask, How do I know whether this is waw conjunctivum or waw
conversivum? This (is how): when before it is another past verb, then it is a
waw conjunctivum; and an example verse is Who has made and done (it)?
[ : qtl.3ms,
: qtl.3ms, Isa 41:4)]. And the wise one will understand. And
likewise the two waws in And one called to another and said [ : qtl.3ms,
: qtl.3ms, Isa 6:3]. Notice, the two are waw conjunctiva because And I saw
the Lord . . . [
: wayy.1s, Isa 6:1] is written before them, which is a past
verb because the waw has a qame. And notice that the style in the Bible is to use
a past in place of a future and a future in place of a past. And this occurs most
often in the words of the prophets, but in historical narrative it occurs very little.
And notice that for second person and first person singular there is another
sign to distinguish waw conjunctivum from waw conversivum: when they have
a waw conjunctivum they generally have a penultimate accent, which is the rule
without the waw, as and I ate the sin-offering today [ : qtl.1s, Lev 10:19]
(and) I spoke by the prophets [ : qtl.1s, Hos 12:11]. They are pasts since
the accent is penultimate but with waw conversivum the accent generally turns to
ultimate as and you will keep all his statutes [
: wqtl.2ms, Exod 15:26],
and I will speak my judgment, etc. [ : wqtl.1s, Jer 1:16]. (my translation)5
5.
. .
.
.
.
.
.
(Leo 1818: 226) .
85
Apart from these various usages, the Future [yiqtol] has yet another, unique and
peculiar to the Hebrews, in that it receives the force of our Past, and designates
a matter as truly past; not however by itself nor absolutely, but viewed in relation to some preceding past event. When different events are to be narrated that
follow the one from the other in some kind of continuous series, the Hebrews
consider the first as past; the others, however, that follow, as future on account of
the preceding. Consequently, this describes something that, in relation to another
past event, is itself later and future; it may be called the Future relativum. (my
translation)6
86
Chapter 2
87
Present
am (Present)
I love/am loving
amv (Perfect)
I have loved
Future
amb (Future)
I shall love
amver (Future Perfect)
I shall have loved
place, Ewald explained, Hence, with reference to action, the speaker views
everything as either already finished, and thus before him, or as unfinished and
non-existent [i.e., not yet existing], but possibly becoming and coming (1879:
1).10 Further on, he wrote: Since, therefore, in virtue of the power and freedom
accorded to the imagination, the ideas of completeness and incompleteness
may also be used relatively, in such a way that the speaker, in whichever of
the three simple divisions of time (past, present, or future) he may conceive
of an action, can represent it either as complete, or as going on and coming
(Ewald 1879: 3).11
The foundation of Ewalds aspectual treatment of the qatal : yiqtol opposition appears to be the Stoic-Varronian tense-aspect theory of Latin (DeCaen
1996: 138). As explained in chap. 1 (1.1), the Stoic-Varronian theory defines
the Latin verb forms according to the parameters of tense and aspect (table
2.2, repeated from table 1.2, p.3). Ewald, however, eschewed the tense parameter since BH has only two primary verb forms; he adopted only the early
aspectual conception of complete and incomplete from the Stoic-Varronian
model as the distinguishing feature in the qatal : yiqtol opposition.
Ewalds adoption of the terms perfectum and imperfectum for qatal and
yiqtol appears to confirm the connection with the Stoic-Varronian theory.
regarded (Ewald 1879: 1, emphasis mine). In light of Ewalds treatment of this opposition,
Kennedys interpretive liberty with the German text appears justified. Nevertheless, DeCaen
(1996: 134) argues that Kennedys use of the term was anachronistic since, as he claims, the
term aspect was first introduced into Western grammar by Georg Curtius (1846) in his study
of the Greek verb. Although the term did not enter the English language until 1853 (so OED,
s.v.aspect), Binnick (1991: 141) notes that Jacob Grimm (17851863) earlier had been
the first to extend the idea of aspect to non-Slavic languages, employing it in his study of
Germanic languages (see above, p.18 n. 15).
10. So fat denn der Redende in Beziehung auf das Handeln alles entweder als schon
vollendet und so vorliegend, oder als unvollendet und nochnichtseiend mglicherweise aber
werdend und kommend auf (Ewald 1870: 349).
11. Da also die Begriffe des Vollendeten und Unvollendeten nach der Kraft und Freiheit der Einbildung auch beziehungsweise (relativ) so gebraucht werden knnen da der
Redende, in welchem der drei reinen Zeitkreise (Vergangenheit, Gegenwart, Zukunft) er eine
Handlung sich denken mag, sie da entweder als vollendet oder als werdend und kommend
sezen [sic] kann (Ewald 1870: 350).
88
Chapter 2
Ewald explained the intended sense of his use of the terms perfectum and
imperfectum: understanding these names, however, not in the narrow sense
attached to them in Latin grammar, but in a quite general way (1879: 3).12
McFall (1982: 44) explains that, by allgemein (general), Ewald presumably
meant their etymological meaning of complete and incomplete, in contrast
to their use as labels for specific Latin morphological forms.13 Thus, Ewalds
theory may be understood as an early aspectual type, in the tradition of the
Stoic understanding of (in)completion. By early, I mean to distance Ewalds
concept of aspect from the more-recent, Slavic-influenced understanding of
aspect that correctly distinguishes perfective and imperfective from the related
notions of complete(d) and incomplete(d) (see 1.3.3).
Although Ewald referred to the aspectual pair in temporal terms, he understood the forms to intersect all three timespast, present, and future: and, on
the ground of this most simple distinction of time a multitude of finer distinctions and forms can be made (Ewald 1879: 3).14 Ewalds taxonomy (1870:
34650; 1879: 37) of qatal includes its use to designate (1) simple past,
(2)past in the past (pluperfect), (3) past in the future (future perfect), (4)pres12. [D]iese Namen aber nicht in dem engen Sinne der Lateinischen Grammatik sondern
ganz allgemein verstanden (Ewald 1870: 350).
13. DeCaen (1996: 13738) disagrees with McFalls interpretation, arguing that, in
fact, there is a second, highly technical, general interpretation of the terms perfectum vs.
imperfectum (though usually the latter is termed the infectum in Latin studies), referring not
to Latins two so-called past tenses, but to the two Latin stems that bear tense inflection.
It is equally clear that Ewald intended this technical, comparative Indo-European, general
sense. This claim, however, is not supported by Ewalds exposition of the BHVS, nor does
DeCaen make any reference to Ewald in his ensuing speculative discussion (DeCaen 1996:
13940). There is no hint in Ewalds grammar that he wants to parallel the BH verb forms
with the Latin stems; rather, DeCaens view relies on reading more into Ewalds statements
than a plain reading (such as McFall has provided) demands.
In truth, DeCaens disagreement is with the aspectual interpretation of the Stoic-Varronian
model of the Latin verb (Robins 1997: 65) more than with Ewalds interpretation of the
BHVS: the imperfectum (infectum) : perfectum distinction in the Stoic-Varronian model
appears to answer to relative tense (DeCaen 1996: 139). This is a gratuitous assumption
not made by other scholars who, nevertheless, disagree with the aspectual interpretation of
the Stoic-Varronian model (see Binnick 1991: 2026).
DeCaens argument is problematic on still other grounds. He (DeCaen 1996: 137) accuses
Ewald of a morphocentric fallacy by excluding the Participle from his theory. However,
DeCaen (1996: 136 n. 20; see above, 1.1) appears to fall into the age-old error of assuming
that the recognition of three times demands that all languages have three corresponding tense
forms. DeCaens accusation (1996: 14041) that Ewald was misled by German romanticism
is likewise unconvincing; if there was an insidious strain in Ewalds treatment of the BHVS
that sought to distance it from the Indo-European verbal system, there must have been a more
successful way of accomplishing this than analogizing the Hebrew verb with Latin!
14.[U]nd auf dem Grunde dieser allereinfachsten Zeitunterscheidung [i.e., vollendet: unvollendet] sind eine menge feinerer unterscheidungen und gebilde mglich (Ewald
1870:350).
89
ent, (5)present perfect, and (6) past conditional. His list of meanings for yiqtol
includes (1) future, (2) present durative, (3) past, (4) present habitual, (5)past
habitual, (6) past conditional, and (7) jussive mood (Ewald 1870: 35058;
1879: 713).
Altogether, Ewald recognized six distinct verb forms or tenses in BH:
two simple tenses (qatal and yiqtol), two modified tenses (wayyiqtol and
wqatal), and two reduced tenses (waw plus qatal and waw plus yiqtol). The
latter reduced tenses are the forms that arose out of the aufgelsten (dissolution) of the modified forms in late BH (Ewald 1870: 842; 1879: 249).
The semantic value of waw in both the modified tenses and the reduced tenses
is that of sequence or consecution of time (1879: 244), including the stronger notion of consequentiality (1870: 84142; 1879: 24748; see Waltke and
OConnor 1990: 477).
Beyond these claims regarding the semantics of the waw-prefixed forms,
Ewald did not greatly advance the understanding of these forms with respect
to qatal and yiqtol. Following the medieval grammarians, Ewald noted the accentual difference between qatal and wqatal in certain forms (Ewald 1870:
600; 1879: 23; see 2.1 and on Levita in 2.2.1 above). With respect to way
yiqtol, Ewald (1879: 19) argued that the form is Jussive based, comparing it
with the past-tense value of the Arabic Jussive preceded by the negative lam
(see W.Wright 1962: 2.41). Ewald theorized that the vocalization of the waw
conjunction on wayyiqtol might be explained by the assimilation of the adverb
( then) between the waw conjunction and verb prefix, and that it is this peculiar element that throws an action into the sphere of the past (Ewald 1879:
19; cf. other theories in McFall 1982: 21719).15 Nevertheless, following his
predecessors, Ewald continued to assume a basic etymological and therefore
semantic identity between wayyiqtol and yiqtol, on the one hand, and wqatal
and qatal, on the other.
At the same time, despite Ewalds efforts to distance his theory from the
tense theories of the BHVS, he could not completely escape the influence of
the relative-tense explanation of the waw-prefixed forms, so popular in his day
(2.2.1). Yet, while he continued to refer to wayyiqtol and wqatal as bezg
licher Zeiten und Modi (relative tenses and moods; Ewald 1870: 593; 1879:
18), Ewalds explanation may perhaps best be described as relative aspect.
But as, in creation, through the continual force of motion and progress, that
which has become, and is, constantly modifies its form for something new; so, in
thought, the new advances which take place (and thus, then) suddenly changes
the action which, taken by itself absolutely, would stand in the perfect, into this
tense, which indicates becomingthe imperfect. . . . As, therefore, in the combination previously explained [i.e., wayyiqtol], the flowing sequence of time or
15.Welches eine Handlung in den Kreis der Vergangenheit verweist (Ewald
1870:593).
90
Chapter 2
thought causes that which has been realized, and exists, to be regarded as passing
over into new realization; so in the present case [i.e., wqatal], it has the effect of
at once representing that which is advancing towards realization, as entering into
full and complete existence. Hence, each of the plain tenses gracefully intersects
the other, by interchanging with its opposite. (Ewald 1879: 20, 2223)16
91
aspect: entering/nascent durative/continuing completed
qatal
Participle
yiqtol
Hebrew conjugation:
Greek stem:
92
Chapter 2
93
attached to a preceding verb by means of this waw consecutive, loses its individuality: no longer maintaining an independent position, it passes under the
sway of the verb to which it is connected (Driver 1998: 118).
2.2.4.Summary
Virtually all subsequent theories about the BHVS have been consciously
presented against the backdrop of the Ewald-Driver standard aspectual theory, which arguably has had a greater influence than any other single theory
about the BHVS. Despite some dissension, Ewalds theory is properly interpreted as aspectual (Fensham 1978; DeCaen 1996); however, his understanding of aspect, like the classical aspectual theories, is expressed in terms of
(in)completion. He popularized the labels perfect and imperfect for qatal and
yiqtol and popularized the term consecutive for the waw-prefixed forms. Driver
retained Ewalds labels and clarified that the BH verbs designate kind of action (i.e., aspect).
Driver, however, departed from Ewald in some significant ways. He rejected parts of Ewalds explanation of the form and etymology of wayyiqtol
and offered nonparallel explanations for the two waw-prefixed forms. Driver
also accorded a greater role to the Participle in his model than previous theories had. This innovation may be seen as sort of a compromise, incorporating
the three tenses of the medieval theories but interpreting them aspectually.
It is perhaps a tribute to the power of Ewalds theory, then, that Drivers work
is often interpreted as simply an English version of Ewalds aspectual theory
Neither Ewald nor Driver, however, was able to progress beyond the early
19th-century explanations of the waw-prefixed forms. Both presumed that the
forms were etymologically and semantically related to their non-waw-prefixed
formal counterparts, and despite their aspectual treatment of the non-wawprefixed forms, they fell back on explanations similar to the relative tense and
waw inductive theories from the early part of their century in order to explain
the semantics of wayyiqtol and wqatal.
2.3. The Biblical Hebrew Verbal System in Historical
and Comparative Perspective
Historical and comparative studies of Semitic languages have made some
the most important contributions to the understanding of the BHVS in the 20th
century.19 This fact is due to the rapid advancement of knowledge of Semitic
languages from approximately the mid-19th to the mid-20th century, instigated
specifically by three notable developments. The first was the decipherment of
Akkadian in the 1850s, which instigated a new phase of discussions regarding
the development of the Semitic verbal system. The second was the discovery
19. For an overview of historical and comparative linguistics, see Anttila 1989.
94
Chapter 2
20. The manner of referring to cognate verb forms varies from language to language
with respect to both the root used in paradigms and the semantic labels associated with the
conjugations. I employ the following conventions in this discussion in an attempt to avoid
confusion. When referring to cognate verbal conjugations, I use the traditional root QTL to
kill in the 3ms form of the base or G-stem (without any * to designate it as a reconstructed
form), ignoring theme vowel variations except where relevant to the immediate discussion;
for example, qatala stands for West Semitic qatala~qatila~qatula, and yaqtulu stands for
Central Semitic yaqtulu~yaqtilu~yiqtalu; yaqtul(u), with the final vowel in parentheses representing both long and short prefix patterns, which were not clearly distinguished in early
discussions. When referring to reflexes of cognate verb forms in specific languages, I use
a root that is employed as the paradigm form in one or another of the standard grammars
of that language; namely, Akkadian PRS, Ethiopic NGR, Arabic KTB, Ugaritic MLK. (This
approach, however, is not followed for BH, for which qatal, yiqtol, wqatal, and wayyiqtol
are employed, as explained above, p.78 n.1. I generally avoid employing the semantic labels associated with language-specific forms, because they can be misleading, and there are
often multiple labels employed in the literature; for example, Akkadian Durative or Present,
Preterit or Past, Permansive or Stative. When I do employ semantic labels, I capitalize them
to designate that they are language-specific labels and not notional or typological categories
(e.g., Durative vs. durative). For full paradigms of the verb conjugations under discussion,
see Moscati 1980: 137, 142; Bergstrsser 1983: 22535; Bennett 1998: 94118.
95
SEMITIC
WEST SEMITIC
CENTRAL SEMITIC
(Ugaritic, Aramaic, Canaanite,
Old South Arabian, Arabic)
EAST SEMITIC
(Akkadian, Eblaite)
SOUTH SEMITIC
(Ethiopian, Modern
South Arabian)
96
Chapter 2
97
98
Chapter 2
and a southern Babylonian dialectfor which three broad stages of development may be discerned, as illustrated in table 2.4 (Huehnergard and Woods
2004: 219; Huehnergard 2005: xxiii; see also von Soden 1995: 2; Buccellati
1996: 12; 1997: 69; 2003: 57).
In addition, dialects of Akkadian that vary to one degree or another from native Mesopotamian Akkadian are described as Peripheral Akkadian. In this
category belong, among others, Amurru Akkadian (Izre'el and Singer 1991),
el-Amarna Akkadian (Bhl 1909; Ebeling 1910; Dhorme 191314; Knudtzon
1915; Moran 2003; Rainey 1996), and the Akkadian of Ugarit (Huehnergard
1989).30 Grammatical descriptions have focused primarily on the Old Babylonian dialect, viewed by many as the classical stage of the language (Buccellati 1996; 1997: 69; Huehnergard 2005: xxvi).
The verbal system of Akkadian features three indicative prefix verb forms.
The iparras form, which is distinguished by the doubling of its middle root
consonant, is described as either Durative (i.e., imperfective) or Present, designating non-past tense (von Soden 1995: 78; Buccellati 2003: 58; Huehnergard and Woods 2004: 253; Kouwenberg 2010: chap.4). Akkadian iparras has
cognates in South Semitic (e.g., Ethiopic ynaggr) but lacks any cognate form
in the CS languages;31 however, its range of meaning is quite close to BH yiqtol
(see Buccellati 1996: 100105; Huehnergard 2005: 9899; Kouwenberg 2010:
91). The past-perfective iprus prefix conjugation is appropriately labeled Preterite or Past (von Soden 1995: 79; Buccellati 1996: 100105; Huehnergard
2005: 1819; Kouwenberg 2010: 12632). Although past-perfective in WS is
predominately expressed by the suffix conjugation qatala, the productivity of
30. For studies of Peripheral Akkadian dialects, see the bibliographical notes in Izre'el
and Singer (1991: 14) and Huehnergard (1989: 7 n. 6).
31.Attempts to find remnants of a yaqattal conjugation in CS Ugaritic and Hebrew
have failed. The main advocate of a yaqattal reflex in Ugaritic was Goetze (1938; 1941ab),
whose arguments have been refuted by Ginsberg (1939) and Fenton (1970). Several scholars
have attempted to find remnants of yaqattal in Hebrew, including Meyer (1958), Rssler
(1961; 1962), Rosn (1969), and Siedl (1971); Bloch (1963) and Fitzgerald (1972) refuted
Rsslers arguments in particular. On the yaqattal conjugation in Afro-Asiatic and Semitic,
respectively, see Greenberg (1952) and Janssens (1972). Rainey (1975: 423; 1990: 41213)
has argued that the Amarna Letters provide no evidence for yaqattal in WS.
99
100
Chapter 2
101
In the West-Semitic languages, however, the form qatala underwent a further development, by which the whole of the tense relations were completely reshaped.
Here, namely, the perfect meaning of the nominal form qatal, such as in words
like culprit, winner, murderer, has made a break through and has been spread
almost to the whole verbal system. Thereby, the form likewise became suited to
serve as a narrative tense, in the way the Perfect of our languages does [i.e., the
European languages]; yaqtul consequently was relieved of its narrative function and was limited to its other not precisely circumscribed uses, which we can
characterize as approximately corresponding to a present participle. Both verbal
forms, as one can see, have nearly exchanged their roles compared with ProtoSemitic. (Bauer 1910: 18; my translation)34
According to Bauer, this semantic role-reversal of the suffix and prefix forms
in WS did not completely obviate their earlier meanings, thus resulting in a
system with two new verb forms alongside the old forms. Bauer argued
that these morphologically identical pairs were differentiated by word stress
and, within BH in particular, by the requisite prefixed waw on the older forms,
as shown in table 2.5; thus, Bauer (1910: 10) suggested that, instead of the
conventional term waw conversivum, the conjunction on these older verb
forms in BH should be termed waw conservativum.
However, Bauer observed that the distinguishing waw is sometimes absent
from qatal in generic (proverbial) sentences, subordinate sentences, and in the
poetic representation of future events (i.e., the prophetic perfect), and from yiqtol in poetry and following certain particles: then, before, until
34. In den westsemitischen Sprachen hingegen hat die Form qatala eine Weiterentwickelung [sic] erfahren, wodurch die gesamten Tempusverhltnisse vllig umgestaltet wurden.
Hier ist nmlich die in Wrtern wie Tter, Sieger, Mrder liegende perfektische Bedeutung der Nominalform qatal zum Durchbruch gekommen und fast auf den ganzen Verbalbe
stand bertragen worden. Dadurch wurde die Form in derselben Weise geeignet, als Tempus
der Erzhlung zu dienen, wie das Perfektum unserer Sprachen [i.e., the European languages];
jaqtul ward infolgedessen seiner erzhlenden Funktion enthoben und auf seine sonstigen,
nicht genau umschriebenen Verwendungen eingeschrnkt, die wir annhernd als die einem
Participium praesentis entsprechenden bezeichnen knnen. Die beiden Verbalformen haben,
wie man sieht, gegenber dem Ursemitischen ihre Rollen nahezu vertauscht.
102
Chapter 2
103
104
Chapter 2
105
106
Chapter 2
and epistolary (Pardee 2004b: 289; Bordreuil and Pardee 2009: 120). The
texts mostly originated in Ugarit and date to the period of the Late Bronze Age
between 1300 and 1190 b.c.e.; the city was destroyed and abandoned in the
early 12th century b.c.e. In addition to Ugaritic, texts have been unearthed at
Ras Shamra written in Akkadian, the international lingua franca of the period
(about 2,000 texts), as well as Sumerian, Egyptian, Luwian, Hittite, Hurrian,
and Cypro-Minoan (Pardee 2004b: 289; Bordreuil and Pardee 2009: 8).
While Ugaritic is recognized as a NWS language (e.g., Ugaritic features
word-initial /w/ > /y/, which separates NWS from CS Arabic), debate has
persisted since its discovery regarding its status within NWS, narrowing to the
question whether Ugaritic is Canaanite or a sister NWS language of Canaanite
(see fig.2.2 above; C. H. Gordon 1965: 14448; see esp.Goetze 1941b; more
recently, cf. Tropper 1994 and Sivan 2000). While Ugaritic shares certain isoglosses that place it closer to Canaanite than Aramaic, the other major branch
of NWS (e.g., // > //; cf. Aramaic // > /q/ or //), it lacks certain earmarks of
Canaanite, including the Canaanite shift (i.e., // > //) and a causative h-stem
or y-stem (cf. Ugaritic causative -stem; see Goetze 1941b; Segert 1984: 1314;
Sivan 1997: 23). Literarily, Ugaritic poetry strongly resembles biblical poetry
formally (parallelism), lexically, and thematically (Gordon 1965: 145; Pardee
2004b: 288). Some of the difficulty in comparing Ugaritic and Canaanite arise
from the chronological difference between the second-millennium Ugaritic
material and the first-millennium Canaanite texts. Pardee (1997: 131; 2004b:
288) observes that Ugaritic has features characteristic of old Canaanite and it
may be a remnant of a Western Amorite dialect (similarly Goetze 1941b).
The Ugaritic writing system is consonantal with the crucially important
exceptions of the three alep signs a, , and u. These three signs are syllabic,
representing a glottal (alep) followed by an a, i, and u vowel, respectively;
syllable-final alep is represented by the sign (see Segert 1984: 2223; Sivan
1997: 910; Tropper 2000: 3339; Pardee 1997: 132; Bordreuil and Pardee
2009: 22). Another important source of information for the vocalization of
Ugaritic are texts with Ugaritic words written in Akkadian syllabary (Sivan
1984; Huehnergard 1987b). Together, these sources form the most direct and
important knowledge about the vocalization of Ugaritic.
The Ugaritic verbal system manifests an archaic WS verb system (Pardee
2004b: 302) consisting of two basic patterns. One is the suffix conjugation
malaka, labled Perfect. Internal evidence exists only for a malika pattern, i.e.,
lk = /laika/ he sent (see C. H. Gordon 1965: 69; Sivan 1997: 113; Tropper
2000: 46971), but Ugaritic texts written in Akkadian syllabary attest to some
malaka forms, e.g., a-ma-ta = /amata/ (Huehnergard 1987b: 31920). The
second, a prefix pattern, is the basis for several conjugations, distinguished
by suffixation. As with malaka, there is evidence for internal ablaut distinguishing dynamic yamluku from stative yamliku and yimlaku patterns: e.g.,
107
amlk /amluku/ I will reign; lak /ilaku/ I will send; syllabic [i]a-ab-i-ru
/yabiru/ (C. H. Gordon 1965: 71; Sivan 1997: 11517; Tropper 2000: 44754;
Huehnergard 1987b: 219 n. 87). The Imperfect yamluku and Perfect malaka
stand in binary opposition as the only two indicative verb forms in prose texts.
The nonindicative (modal) system features an Imperative, built on the prefix pattern base, *m(u)luk; a Jussive yamluk, which is distinguished from the
Imperfect by a zero suffix; a Subjunctive or Volitive a-suffixed yamluka; and
two energic forms with suffixed -n(n) (Energic 1 yaqtulan and Energic 2 yaqtulanna), the meanings of which are not entirely clear but are generally associated with the modal system (see Pardee 2004b: 3036; C. H. Gordon 1965:
7173; Segert 1984: 5663; Sivan 1997: 11519; Tropper 2000: 497506).
Despite some dissenting opinions (esp. Greenstein 1998; 2006), it is widely
recognized that Ugaritic had a reflex of PS prefixed past yaqtul, homophonous
with the Jussive yamluk (Lipiski 2001: 347; Sivan 1997: 98100; Tropper
2000: 683, 69597; Pardee 2004b: 303; but cf. Bordreuil and Pardee 2009: 46).
Crucial to the semantic description of the Ugaritic verbal system is the
prose: poetry distinction. In prose, the verbal conjugations function analgously
to their cognate forms (e.g., BH, Aramaic, Arabic). However, the distribution
of the suffix and prefix conjugations in Ugaritic poetic texts has thus far defied
complete description (Pardee 2004b: 303). Pardee has criticized attempts to
describe the Ugaritic verbal system on the basis of the poetic texts or to theorize a system that is equally applicable to prose and poetry, surmising that the
poetic texts reflect an older stage of the language when the zero-ending imperfect form functioned as a preterite, like the Akkadian iprus (Pardee 2004b:
303; 20034: 14; 1997: 13).
2.3.3.2. The Northwest Semitic Verbal System in Light of Ugaritic
Granting the singular importance of the Ugaritic evidence to our knowledge
of NWS, we nevertheless find that the contribution of the data on its verbal
system is particularly difficult to assess because of several complicating factors.38 The first is simply the nature of Ugaritic philology: the texts are mostly
consonantal, often fragmentary, and many still await authoritative publication
(Pardee 2003-4: 12). Second, and following naturally on the first point, debate continues unabated regarding the semantics and functions of the Ugaritic
verbal forms (e.g., Richardson 1991; Tropper 1991; 1992; 1995; Rainey 1990;
M.S. Smith 1995; Sivan 1998; Greenstein 1998: 40913; 2006). Finally, based
on continued debate over the genetic affinities of Ugaritic, scholars interpretation of its verbal system has often been via the paradigm of related Semitic
languages, particularly Hebrew. A case in point is C.H. Gordons comment
38. Emerton (1994b: 54) comments, perhaps somewhat tongue-in-cheek, on the contribution of Ugaritic to our understanding of Hebrew: Whether it [Ugaritic] has shed more
light or raised more problems is a debatable question.
108
Chapter 2
(1965: 6869) that in fact yqtl is the regular narrative form and we shall often translate it as a historical present. Richardson (1991: 285) surmises that
Gordons puzzling treatment of the form as a historical present stems from his
underlying equation of Ugaritic yamluk(u) and BH yiqtol.
However, Hebrew was not the first framework through which the Ugaritic
verbal system was viewed. Goetze, in one of the earliest descriptions of the
Ugaritic verbal system, approached it from Akkadian. Goetze (1938: 286) argued that the Ugaritic suffix conjugation should be vocalized malika, similar
to the Akkadian counterpart paris/parsku, and that it predominately has a
stative connotation and is therefore not an integral part of the system of
tenses (Goetze 1938: 289). Goetze (1938: 28996) identified three different
prefix conjugations in Ugaritic: the indicative yamluku, regularly used as a
narrative past in poetic texts; the nonindicative subjunctive yamluka; and the
apocopated jussive yamluk. On the basis of his semantic description of these
three conjugations, Goetze (1938: 296) argued that, because none of the three
properly expresses present tense, a present tense yaqattal conjugation had been
and should be surmised for Ugaritic.
Through a series of exchanges between several scholars in the 1960s
1970s, claims to have found reflexes of PS yaqattal in Ugaritic and other NWS
languages were discredited (see p. 98 n. 30 above). Fenton, who argued
forcefully against Goetzes view in this matter (Fenton 1970), subsequently
interpreted the BHVS in light of Ugaritic (Fenton 1973). Fenton (1973: 34)
stated that Hebrew and Ugaritic belonged to the same linguistic continuum
and their tense systems shared the same history and drew appropriate lines of
developmental connection between the two. In poetry, the BHVS, not surprisingly, appeared to be at a later stage of development than the Ugaritic verbal
system in that qatala was used more freely as a narrative form alongside the
omnitemporal yaqtul(u) form. By contrast, in prose, the BHVS appears to be
the more conservative system since it preserves the yaqtul preterite in wayyiqtol in contrast to Ugaritic, in which qatala is the exclusive form for recording
single past events (Fenton 1973: 34). Similarly, Held (1962: 282) saw clear
parallels in the Ugaritic and BH verbal systems in poetry, except that he argued
that, when yaqtul(u) appeared parallel with qatala in BH or Ugaritic, it was not
an omnitemporal verb form but a reflex of the PS Past yaqtul.
Fenton (1973: 32) found in the Ugaritic data confirmation of Bauers (and
G.R. Drivers) view of the wayyiqtol in BH as derived from an earlier/eastern
past yaqtul. By contrast and earlier, Harris (1939: 11) had claimed that the
Ugaritic evidence countered Bauers theory in that it showed that the BHVS
was no Mischsprache of ES and WS dialects but was a native NWS verbal system.39 In this regard, Ugaritic has provided invaluable data on the early NWS
39.Although Harriss work is entitled Development of the Canaanite Dialects, it treats
Ugaritic on the supposition that it is part of Canaanite; thus his treatment is essentially of
NWS languages exclusive of Aramaic (Harris 1939: 1011).
109
verbal system: it has confirmed the presence of a short final vowel on qatala
(like Arabic) in contrast to ES Akkadian paris/parsku, and it establishes the
form of the prefix conjugations, the Indicative yaqtulu, Jussive yaqtul, and
Subjunctive yaqtula (Harris 1939: 78; see Bauer and Leander 1991: 296300;
3078). Ugaritic also appears to support the identification of a past prefix form
in NWS, though the Ugaritic data led Harris (1939: 4648) to identify both a
yaqtulu preterite and a yaqtul preterite. Harris claimed that the outstanding
change in morphological structure of Canaanite [i.e., NWS minus Aramaic]
was the replacing of the objective aspect [i.e., tense] system by a subjective
aspect [i.e., viewpoint aspect] system through the following developments:
As the perfect [qatala] replaced the preterite [yaqtul(u)] in regular speech, the
few cases of the preterite which still occurred (in old stereotyped phrases and the
like) came to be formulaic, members of an obsolescent form, and were no longer
understood as a preterite tense. Now it happened that this petrified verbal form,
yaqtulu, seemed to have an obvious relation to the other existing forms: as between yaqatalu present indicative,40 and yaqtul, yaqtula jussive and subjunctive
aspects, this yaqtulu had precisely the form which an indicative aspect related
to the two modes would have had. It is probable that on this formal analogy the
petrified yaqtulu forms came to be interpreted as a new imperfect indicative, a
category hitherto non-existent. This new imperfect soon became a regular construction, applicable to any root for which there was a jussive or present. . . .
Fitting as it did into the now largely subjective system of the Canaanite verb,
it was favored over the present. In time, the present dropped out of use and Canaanite was left with a practically pure subjective-aspect verb system, the chief
remnant of the old tense system being the short preterite in the inland dialects.
(Harris 1939: 8485)
While Harriss early understanding of TAM in the NWS verbal system is understandably no longer credible, TAM in Ugaritic remains unsettled, judging
from the grammars. Gordon apparently treated Ugaritic malaka and yamluku
aspectually, as equivalent to their BH cognates, though he assiduously avoided
the semantic labels perfect and imperfect (C. H. Gordon 1965: 6869;
see Richardson 1991: 28485). Segert (1984: 8889) explained that Ugaritic
opposition malaka : yamluku expressed a perfective : imperfective (constative: cursive) opposition at its early strage; these meanings are preserved in
poetry, where the forms may function in all three temporal spheres. By contrast, in prose texts the forms have temporal character. Sivan (1997: 96103;
1998), following his teacher Rainey, treats the Ugaritic verbal system in terms
of tense: malaka expresses past tense, while yamluku expresses present-future
tense; unfortunately, these tense identifications appear prima facie at odds with
his taxonomies of the forms as used alike in all three temporal spheres. Troppers recent grammar (2000) returns to a more traditional aspectual approach
40. Harris (1939: 49) accepted Goetzes argument (1938) for a reflex of Proto-Semitic
yaqattal in Ugaritic.
110
Chapter 2
to the forms, albeit recognizing also past tense yaqtul (versus imperfective
yaqtulu) in the system (Tropper 2000: 68284; see also 1992; 1995; 1999:
109). Finally, Greenstein (2006) has recently argued against claims of a prefixed past form in Ugaritic and, further, that the default meanings that malaka
and yaqtulu have in prose, which he identifies as past and present-future, respectively, are in practice subordinated to all manner of discourse and rhetorical factors (Greenstein 2006: 78; and Bordreuil and Pardee 2009: 46, who
endorse his position). The implication appears to be that it is still impossible to
construct a coherent semantic analysis of the Ugaritic verbal system in poetry
(see also Greenstein 1988; 1998: 40913).
2.3.4. Amarna and the Canaanite Verbal System
One of the most important contributions that Ugaritic studies have made to
understanding the BHVS is indirect: although discovered in 1887, the Canaanite influences on the Akkadian language of the Amarna correspondences41 were
not fully appreciated until they were viewed in the light of their close contemporary, Ugaritic (see Rainey 1996b: 4; Moran 2003: 6). The Amarna Letters
were discovered at el-Amarna, the capital and residence of the Egyptian king
Akhenaten (Amenhotep IV; ca. 13531336 b.c.e.) and contain diplomatic correspondence from Mesopotamia, Syria, and Asia Minor. Several excavations
uncovered over 350 texts (see Moran 2003: 22325, 23741, 34344). While
the letters are almost entirely written in Babyonian, the lingua franca of the
period (exceptions are some Assyrian, Hurrian, and Hittite letters), they are not
written in standard Babylonian but in Akkadian dialects reflective of the native
languages of the regions from which the letters originated (Moran 2003: 343).
The dialects of Amarna Akkadian along with the dialects of other Akkadian
texts found at Ugarit (Ras Shamra, Syria), Emar (Tell Meskeneh, Syria), Nuzi
(administrative center of the Mitanni empire; modern Yorgan Tepe, Iraq), Alalakh (Tell Atchana, Turkey), Hattusas (the capital of the Hittite empire; modern
Boazky, Turkey), and other places are collectively classified as peripheral
Akkadian (Moran 2003: 237; Huehnergard 2005: xxv).
41. In this discussion, Canaanite refers to the Semitic dialects spoken in the geographical region of ancient Canaan in the second half of the second millenium b.c.e., in contrast to
its more conventional reference to Canaanite dialects of the first millenium as distinct from
other NWS dialects (Pardee 2004a: 386). Moran (2003) and Rainey (1996) use the term with
this former meaning, although Rainey (e.g., 1996b: 221) freely mixes the term with West
Semitic and Northwest Semitic. The Canaanite of Amarna period is clearly CS, based on
the presence of the innovative yaqtulu verb form. In the following discussion, WS, CS, and
NWS are used to refer to the domain of a particular feature exhibited in the Canaanite of the
Amarna Letters.
111
112
Chapter 2
113
ing degrees of success, Moran also sought to delineate the semantic range of
each conjugation based on its syntactic distribution. Raineys comprehensive
study of the evidence in the Amarna Letters from Canaan has substantiated and
supplemented Morans earlier findings (Rainey 1996b). In particular, Rainey
has championed the admittedly meager evidence in the Amarna Letters for a
WS prefixed past verb form (i.e., yaqtul; e.g., Rainey 1986; 1990; 2003). The
findings of Moran and Rainey regarding the Canaanite (proto-Hebrew) verb
are summarized here.
Morans study of the Byblian Amarna evidence confirmed the stative: dynamic theme-vowel distinction already recognized in WS qatala (i.e., qati/
ula: qatala) in contrast to the predominance of the qatil theme-vowel pattern
in Akkadian (Moran 2003: 2829; Rainey 1996b: 29596).43 Importantly, the
qatila pattern in the Amarna evidence sometimes expresses a passive sense
with active (transitive) verbs (Moran 2003: 29; Rainey 1996b: 3035), just
as the Akkadian Verbal Adjective paris regularly does (see Buccellati 1988;
1996: 356; Huehnergard 1987a; 2005: 4.3; von Soden 1995: 77); this is illustrated by example [2.10].
[2.10] laqi (la-q-u) ina ugri ade [ana] uruialunaki
a.On a-de4-e uruia-lu-naki, see Moran 1987: 514 nn. 1718; 1992: 328; 2003: 25960.
Moran observed that, in Byblian Canaanite (as in Ugaritic prose), WS qatala had largely taken over the past narrative functions of PS yaqtul (Moran
2003: 30; Rainey 1996b: 365). In addition, qatala appears 33 times with a
future sense in the Byblian data: 24 are preceded by u and, while 8 without u
43. Moran referred to the verb classes as stative and active; Rainey labels them
stative and transitive.
114
Chapter 2
occur in the protasis of a conditional clause (Moran 2003: 31). Rainey supplemented Morans data with similar examples from the other Amarna Letters
from Canaan: qatala has a future sense in conditional protases and apodoses, in
purpose clauses, and when used as a future-optative (Rainey 1996b: 35565).
Moran, followed by Rainey, drew two main conclusions from this evidence:
first, the WS qatala in Amarna Canaanite shows an earlier stage of development than BHa stage at which the stative present meaning of the adjectivalpredicative origin of the form is more prominant, and functions typical of the
Akkadian Verbal Adjective are still associated with the form (Moran 2003:
35, 216; Rainey 1996b: 365); second, the future sense of qatala derives from
its generally restricted conditional context, which may point to an optative
or precative origin for BH wqatal (Moran 2003: 3133, 216; Rainey 1996b:
366). Moran argued that WS qatala should be defined, by virtue of its origin in
a verbless predication, as a tenseless aorista form incompatible with either
tense or aspect (in the Ewald-Driver sense of complete : incomplete; Moran
2003: 3334; Rainey 1996b: 36566).
One of the most significant discoveries in the Byblian Amarna Letters was
a clear morphological and semantic distinction between yaqtulu and yaqtul.
Although the difference had already been posited for BH and NWS on the
basis of internal (e.g., BH vs.
;see Holmstedt 2000) and comparative
evidence (Arabic yaktubu vs. yaktub; Bauer and Leander 1991: 27374), morphological proof of the distinction in NWS only came with the discovery of
Ugaritic, and the semantics of the distinction was clarified only with Morans
study of the forms in Amarna (Moran 2003: 3839). Although the -u suffix appears in the Amarna Letters as the Akkadian subjunctive or subordinate marker
(on the form, see von Soden 1995: 83; Huehnergard 2005: 18384), Moran
demonstrated that the final -u may also indicate the CS yaqtulu conjugation,
which has a similar semantic range as the Akkadian iparras conjugation (Moran 2003: 4147). Raineys study has underscored Morans earlier conclusions
(Rainey 1996b: 195202).
The semantic range of yaqtulu in the Amarna Letters is demonstrated by
example [2.11].
[2.11] miya mr Abd-airta ardi kalbi ar mt kai u ar mt
mitan(n)i unu u tilqna (ti-l-q-na) mt arri ana um
pannu tilqna (ti-[l-q][-n]a) lni aznka u qlta ann
inanna dubbir rbiaka u laq lniu ana unu anumma
laq l ullaza umma kiama qlta adi tilqna (ti-l-q-na) l
umura u tidkna rbia u b tillati a ina umura
Who are the sons of Abd-airta, the slave and dog? Are they
the king of the Kassites or the king of the Mitanni that they
take (yaqtulu) the royal land for themselves? Previously they
used to take (yaqtulu) the cities of your governers, and you
115
116
Chapter 2
117
118
Chapter 2
119
data, though not completely clear, appears to exhibit the form; and the Amarna
Letters shows evidence of the form in second-millennium Canaanite. To these
data may be added others from Amorite onomastica and NWS epigraphs.
Although no Amorite texts have survived, the grammar of Amorite can
be deduced from onomastics found in Akkadian texts (Gelb 1980). Knudsen
(1991: 879) points to the example of the Assyrian king Shamshi-Adad (ca.
1800 b.c.e.), of Amorite stock, who appointed his two sons administrative
posts; the older son, Ime-dagn (Dagan heard), adopted an Akkadian name
with a Past prefix iprus form, whereas the younger son, Yama-hadd (Hadad
heard), took a WS name with a cognate Past prefix form.
Numerous examples of a WS prefixed Past yaqtul form have been discovered in NWS epigraphsrarely without the prefixed waw (e.g., Tel Dan), more
frequently with waw, as in BH wayyiqtolincluding Zakir (or Zakkur), Deir
Alla, Mesha, and Tel Dan, not to mention the examples of the form in the
Hebrew epigraphs (see Garr 2004: 18486; M. S. Smith 1991: 1819).46 Garr
(2004: 186) concludes: [T]his distribution suggests that the consecutive imperfect [i.e., wa(y)yiqtol] was a common NWS verb form.
The impact of historical-comparative investigations on the understanding of
the BH qatal is no less dramatic. The late-19th-century suggestion by Knudtzon that WS qatala (and BH qatal) had developed from a verbless construction
(specifically, the predicative use of a verbal adjective qati/ul form) has received ample support in data from Amarna and Ebla that link the development
of the form to the ES Verbal Adjective paris/parsku. The Amarna Letters
show the WS qatala functioning with semantic similarities both to Akkadian
(e.g., transitive verbs with passive sense; see example [2.10], p.113) and to
later WS texts. In Eblaite, despite its affinities with Akkadian (e.g., yaqtul:
yaqattal opposition), the suffix pattern functions as active dynamic with transitive verbs (e.g., ba-na-a he built; Mller 1984: 15456; Diakonoff 1990: 27;
C. H. Gordon 1997: 110; cf. Huehnergard and Woods 2004: 252).
By comparison with the etymological distinction evidenced in the historicalcomparative data for wayyiqtol: yiqtol (i.e., yaqtul: yaqtulu), no evidence has
been forthcoming to sustain the claims of Bauer and others that qatal and
wqatal are etymologically distinct. Rather, the investigations have pointed
46. Zakir: w ydy and I lifted my hands (A11), wynny and he answered me (A11),
wymr and he said (A15); Deir Alla: wytw and they came (I 1), wymrw and they said
(I 2), wyqm and he arose (I 3), etc.; Mesha: w and I made (3, 9), wybn and he built
(10), and whrg and I killed (11, 16); Tel Dan: without waw: yssq he went up (2), yhk he
went (3); with waw: wykb and he lay down (3), wyl and he entered (3), wyhmlk and he
made king (4), wyhk and he went (5), wpq and he went out (5), wqtl and I killed (6),
wm and I made (9, 12; see Emerton 1994a; 1997; Muraoka 1995; 1998; Mller 1995a;
Sasson 1997; 2001; Tropper 1996); Hebrew epigraphs: wylkw and they came (Siloam4),
wyqr . . . wykl and he harvested and he measured (Yavneh Yam 45), wylhw he brought
up (Lachish 4.67).
120
Chapter 2
121
provided an orientation to the debate. The description of the standard aspectual theory, established by Ewald and mediated to the English-speaking world
by Driver (2.2), provides necessary background to verb theory of the past half
century. Finally, the survey of historical and comparative studies of the Hebrew
verb (2.3) points to one of the most important developments in Hebrew verb
theory of the 20th century. One could justly add alongside this latter development the rise of modern linguistic theory as being just as important (or more
so) to the advancement of Hebrew verb theory. Although space does not permit
a formal introduction to modern linguistics, reference is made in the following
survey to its influence on Hebrew verb theory.
Two primary options present themselves for organizing this survey. One is
a historical survey in which developments in Hebrew verb theory are traced
linearly over the past half century. This is a fascinating story and could fill an
entire volume on its own; and frankly, a survey of this sort is a desideratum.
However, instead, I have organized this survey in terms of families of theories. The most basic categories of Hebrew verb theory are aspectual theories
(2.4.1), tense theories (2.4.2), and discourse theories (2.4.3). In addition,
the last two decades have seen the rise of theories that attempt to do justice to
the insights of each of these other groups of theoriesa point that is treated
in the following section (2.5). Within some of these categories, subdivisions
of theories according to pedigree may be made (e.g., Longacrean discourse
theory vs. the Weinrich-Schneider school). At the same time, my presentation
is strongly historical because there are notable trends with respect to the waning and waxing of popularity of certain types of theories.
This theory families approach may appear reductionistic. After all, many
theories describe the Hebrew verb as a combination of features, and certain
theories escape categorization according to this scheme. This approach is justified, however, in that these are the terms in which the debate is regularly
cast, and it is therefore expedient to summarize the debate within this familiar
framework. At the same time, by qualifying these theory types with the adjective prominent (a term borrowed from Bhat 1999), I hope to soften the rigidness with which these categories are often employed in the debate. Few, if any
single-parameter theories of the Hebrew verb exist; however, generally, one
or the other of these parameters is identified as most prominent in the system.
These shortfalls notwithstanding, this survey presents the lay of the land of
Hebrew verb theorya worthy goal in itself and a necessary prerequisite for
the presentation of my theory of the Hebrew verb in the following chapters.
2.4.1. Aspect-Prominent Theory
Aspect-prominent theories are heirs of the Ewald-Driver standard theory
in one form or another. They recognize that the most basic distinctions in the
Hebrew verbal system are aspectual. Despite Bauers concomitant rejection
122
Chapter 2
of aspect for tense distinctions in his diachronic analysis of the Semitic verb
(Bauer 1910; see 2.3.2.1), it did not take long for scholars to recognize that
Bauers historical conclusions were fully compatible with the standard aspectual analysis. For example, while accepting Bauers contention that the
waw-prefixed forms were archaisms (Cohen 1924: 19), Cohen argued, contra
Bauer, that the forms were aspectual: the imparfait (yiqtol) and le parfait
en rle dimparfait (wqatal) are linaccompli, while the parfait (qatal)
and le imparfait en rle de parfait (wayyiqtol) are l-accompli (Cohen 1924:
1012, 286). At the same time, a new level of linguistic sophistication can
be seen in Cohens theory, which is evident in the title of his monograph, Le
systme verbal smitique et lexpression du temps: although the formal system
is defined by aspect, he recognized that the verbal forms express a variety of
temporal nuances.47
A renewed interest in the aspectual analysis of the BHVS appears in the
1950s and 1960s, the beginning of which is marked by Brockelmanns 1951
article, in which he defined the Semitic verb as expressing subjektiven Aspect (Brockelmann 1951:1 34; see also 1956: 39; see comments in Mettinger
1974: 65). Significantly, this article represented a reversal of his earlier analysis of the Semitic verbal systems as expressing tense (Zeitstufen; Brockel
mann 190813: 14451). Equally significant was Brockelmanns departure
from Ewalds perfectum: imperfectum terminology in his introduction of the
alternative Latinate terms konstatierend Aspekt (Lat. constare stand still,
exist) and kursiv Aspekt (Lat. cursus running, coursing; Brockelmann
1951: 146; see also 1956: 39). This terminology was subsequently adopted by
other European scholars, notably Rundgren (1961; 1963) and Meyer (1960;
1964; 1966; 1992).
2.4.1.1. Rundgrens Privative Opposition Theory
Rundgrens theory is constructed on the linguistic principles of synchrony
and privative oppositions. Although his theory draws on historical-comparative
research, he chose to treat the BH data as deriving from a single synchronic
entity (Mettinger 1974: 74). The concept of privative oppositions derives from
phonological theory in the Prague school of linguistics (see p.34 n.28). The
concept can be illustrated in the realm of semantics by the example of English
bitch and dog, which are distinguished by the feature of [+female], which the
former term has and the latter lacks. Dog can express two values with respect
to [+female] bitch, to which it stands in privative opposition: (1)it may express the negative counterpart of male canine [+male], or (2)it may express
the neutral meaning of canine [female] or [male] (a similar example is the
opposition drake: duck).
47. On the linguistic distinction between form and substance (meaning), see de Saussure
1959: 11122; Lyons 1968: 5470.
123
present-future B1 Present
time
yiqtol
past time
1 Coincidental
(qatal)
B2 Imperfect 2 Punctual
yiqtol (long) Aorist
yiqtol (short)
124
Chapter 2
105] that qatal early on might have had the same punctual past meaning as
preterite [way]yiqtol) and at the BH stage (i.e., modal wqatal).
Rundgrens model has been praised as a convincing aspectual model of Hebrew (Mettinger 1974: 74), and it has remained current in Scandinavian studies
of BH, being adopted by his students in their studies of related aspects of Hebrew grammar (Isaksson 1987; Eskhult 1990). However, Rundgrens model is
problematic, most of all because qatal does not always or even mostly express
stative eventsa fact that Rundgren admits but explains as being due to the
encroachment (of sorts) of qatal into semantic domains that are dominated by
the dynamic prefix conjugation(s).48 In addition, Rundgrens model is idealized to the degree that he must force the BH verbs into categories that are
not clearly distinguished by the data. For example, it is not clear what the
semantic difference is between Punctual yiqtol and Neutral wayyiqtol.
2.4.1.2. Meyers Diachronic Systemberlagerung Theory
In contrast to Rundgrens deductive, synchronic treatment of the Hebrew
verbal system, the contemporary theory by Meyer is preeminently determined
by the historical and comparative data. Meyers theory is largely an attempt
to resolve the data from Afro-Asiatic and Semitic (particularly East Semitic)
comparative investigations (notably Rssler 1950, ET 1981) and Ugaritic and
Amarna Canaanite studies (see 2.3 above).
Based on the evidence of a Semitic-wide prefix-form opposition of yaqtul:
yaqattal (see Rssler 1950) and the narrative use of yamluku in Ugaritic epic,
Meyer reconstructed the Old Canaanite (including Ugaritic) verbal system as
consisting of five prefix forms based on the Common Semitic yaqtul/qtul (i.e.,
Preterite/Jussive yaqtul and Imperative/Infinitive qtul; Meyer 1960: 31013;
1992: 96103):
1. Preterite/Jussive yaqtul (basis of the Imperfect Consecutive wayyiqtol form)
2. Narrative (Indicative) yaqtulu (based on yamluku in Ugaritic epic narrative)
3. Durative yaqattal (based on evidence of its widespread presence in Semitic)
4. Finalis (subjunctive) yaqtula
5. Energic yaqtulan(n)
The central opposition in this system was between narrative (indicative) yaqtulu and durative yaqattalu, which Meyer posited on the basis of Rsslers
work (1950), in particular, and which he described with Brockelmanns terminology (1951; see 2.4.1 above) as konstatierenden Aspekt versus kursiven
Aspekt (Meyer 1960: 31112). Meyers understanding of the BHVS centers
on his understanding of the direct and concomitant effects of the development
48. The effect of seeing the stative as not merely earlier but the more basic meaning
for qatal is evident in Isakssons study (1987: 7592) of the form in Qoheleth, about which
Schoors (1992: 174) criticizes him for his tendency to multiply the instances of a perfect
tense with a present force (see Cook forthcoming).
125
of fientive (dynamic) qatala in West Semitic from the Common Semitic stative
qatila form against the background of this Old Canaanite verbal systemthe
overlay of a younger system on the older one (Meyer 1960: 31314). This
mile-stone development (so Mettinger 1974: 71) precipitated three effects in
the Canaanite verbal system. First, the new qatala form appropriated some of
the past-narrative (prterital-erzhlende) and jussive functions of the Preterite/Jussive yaqtul, as well as the narrative-past (erzhlend-prterital) function of the narrative (indicative) yaqtulu (Meyer 1960: 314). Later on, in the
form of the Perfect Consecutive, it also assumed present-future functions
(Meyer 1960: 316). Second, concurrently the Durative yaqattal fell into disuse,
due in part to its formal similarity with the Semitic (Intensiv) D-stem (1960:
315). Third, the yaqtulu was in turn confined primarily to present-future functions and fell together with the Finalis yaqtula when final short vowels were
elided throughout the system (Meyer 1960: 316).
Although Meyers view of the BHVS is superficially similar to Bauers
Mischesprache theory, Meyer appears to view the development of qatala as
an unwelcome intrusion into the Old Canaanite verbal system. Meyers theory
also has several obvious problems. For one, the idea that Canaanite ever possessed a reflex of the yaqattal form is now thoroughly discredited (see p.98
n.31 above) but is a major component of Meyers Old Canaanite verbal system (and a feature for whose neglect he criticized Rundgrens theory; Meyer
1964: 123). The other difficulty with Meyers theory is what it leaves unexplained. For instance, what is the relationship between the past-narrative yaqtul
and narrative-past yaqtulu meanings? It seems clear in hindsight that Meyers
theory, in that it attemps to give due attention to the Ugaritic evidence and the
comparative Semitic data, is forced. The difficulty seems primarily to lie in
his uncritical acceptance of how yamluku functions in Ugaritic epic narrative,
given how unsettled this issue remains (see Greenstein 2006).
Mllers more recent work on the Semitic and BH verb represents a continuation (with certain revisions) of Meyers theory (Mller 1983; 1986; 1991;
1998). Mller, like Meyer, takes a diachronic approach, explaining the BHVS
in terms of the effects of a Systemreduktion in the collapsing of the distinction
beteween yaqtul and yaqtulu, and a Systemberlagerung in the development
of qatal (Mller 1983: 56; 1998: 151). However, Mller remedies the major
weaknesses in Meyers theory by (a)downplaying the presence of the yaqattal
conjugation in West Semitic (1983: 44; 1986: 373; 1998: 150) and (b)shifting
the center of his theory to the West Semitic opposition between the short and
long prefix forms, yaqtul: yaqtulu: the former of these had a preterite-narrative
function (> BH wayyiqtol) as well as a jussive meaning, while the latter expressed imperfective or durative aspect and present-future tense (Mller 1983:
3738; 1986: 37071; 1998: 14749).
126
Chapter 2
Mller explains that the Hebrew verbal system moved toward a purely tensed
system in the late and postbiblical periods, creating a three-conjugation tense
system of qatal (past), Participle (present), and yiqtol (future) (Mller 1998:
15051).
Two peculiarities may be noted about Mllers treatment of the BHVS.
The first is his denial of a semantic distinction between the waC- and wmorphemes (Mller 1991), emphasizing that the morphological distinction between wayyiqtol and wyiqtol reflects the historical opposition between short
yaqtul and long yaqtulu, respectively (Mller 1983: 40). The other peculiarity
is his division between subjective Aspekt and objective Aktionsart, reflected
in the distinction between perfective and momentary-punctual and between
imperfective and durative. Yet these terms are often used interchangeably, thus
making it difficulty to determine the significance of Mllers distinction between these categories (see Binnick 1991: 13949).
2.4.1.3. Michels Synchronic Theory
Michel approached the BHVS with an inductive, synchronic approach, in
conscious opposition to historical and comparative theories. He objected that
comparative linguistics is only possible if the languages being compared are
understood in their own right (Michel 1960: 14). Examining the verbal system
in the book of Psalms, and rejecting diachronic evidence, Michel concluded
that there was virtually no distinction between the waw-prefixed forms and
their non-waw-prefixed counterparts: qatal (including wqatal) represents
situations as selbstgewichtig, whereas yiqtol (including wayyiqtol) presents
situations as relativ to some other event (Michel 1960: 254). While he does
not recongize any semantic difference between yiqtol and wayyiqtol, based on
his close attention to syntax Michel claims that wayyiqtol represents a situation
that is in closer relation to what precedes than yiqtol (Michel 1960: 47, 132).
In addition, Michel distinguishes the prefix pair and the suffix pair in terms of
how they relate their subject to the action: with the qatal (and wqatal), the
action has an akzidentiellen Charackter with regard to the subject, whereas
127
the yiqtol (and wayyiqtol) has a substantiellen Charakter (Michel 1960: 110,
127).
Although Michels explanation appears prima facie to be quite different
from anything preceding it, it does have points of connection. On the one hand,
Michel uses terms remincent of earlier aspectual treatments when he describes
the qatal form in isolation or at the beginning of a sentence as expressing
konstatiert es ein Faktum (Michel 1960: 47). On the other hand, Waltke and
OConnor (1990: 473) point out the similarity between Michels description of
the Hebrew verb opposition and Turners 19th-century theory: 49
It might be said that the first [qatal] is the more abstract, the second [yiqtol] the
more concrete,the one the more objective, the other the more subjective. . . .
Perhaps the most proper words which our language affords for the expression
of the distinction are these,the Factual and the Descriptive. The one makes
statements, the other draws pictures; the one asserts, the other represents; the one
lays down positions, the other describes events; the one appeals to reason, the
other to imagination; the one is annalistic, the other fully and properly historical.
(Turner 1876: 384)50
49. Waltke and OConnor (1990: 47475) also note the similarity of Kustrs conclusions to Michels, though Kustr uses different terminology, describing the opposition
between qatal (including wqatal) and yiqtol (including wayyiqtol) in terms of determining (determinierend) and determined (determiniert) action, respectively (Kustr 1972:
4546, 55).
50. McFall (1982: 77) offers the following succinct description:
The essence of Turners theory is that qtl expresses the action or state as the attribute
of the person or thing spoken of; the yqtl form expresses or represents the verbal action as in or of the subject, the produce of the subjects energy, the manifestation of its
power and life, like a stream evolving itself from its source. Whereas the first represents
the act or state as an independent thing: the Factual; the second expresses the same
act or state as a process, and one that is passing before our very eyes: the Descriptive.
128
Chapter 2
is no semantic difference between the the waw-prefixed forms and their formal
non-waw-prefixed counterparts.51
The most recent permutation of this type of theory comes from Furuli
(2006), who claims based on his examination of all the verb forms in the Hebrew Bible, Ben Sira, and the Dead Sea Scrolls that there are only two conjugations: a perfective suffix conjugation (qatal and wqatal) and an imperfective
prefix conjugation (yiqtol and wayyiqtol; Furuli 2006: 4). He concludes that a
diachronic approach is problematic and unnecessary, since (1)there is no evidence of a prefix preterite in Northwest Semitic or Akkadian, and (2)there is
no evidence in his corpus of any semantic change in the Hebrew verbal forms
(Furuli 2006: 147).
Furulis approach is based on two premises that he claims are not taken into
account by any earlier theories. This first is the systematic distinction between
past time reference and past tense. Although it is a sound principle, Furuli uses
this premise to dismiss out of hand any and every tense explanation of Hebrew
and Semitic (e.g., Furuli 2006: 32, 98). Furuli claims that context can fix the
temporal reference of a verb and then refuses to acknowledge any other possible means of fixing temporal referencethat is, tense (Furuli 2006: 100).
His second principle is that aspect (viewpoint aspect, in particular) in Hebrew is of a different sort than English aspect, which he claims informs most
previous understandings. Unfortunately it is unclear to me the basis for his
claim because the only explanation he offers is, because aspect is a kind of
viewpoint, it is not obvious that it has the same nature in the different aspectual
languages of the world (Furuli 2006: 49).
A full survey of Furulis work would take too long and yield too little of
value (see the reviews of Kummerow 2007 and Cook 2010). Here I mention
the two fundamental difficulties with his theory that are most germane to this
discussion: his treatment of wayyiqtol and his analysis of aspect. A major if
not the central focus of Furulis work is to show that wayyiqtol is not a distinct
prefix form from yiqtol (and wyiqtol) but is an invention of the Masoretes. He
recognizes that a major obstacle to his argument is that 93.1% (according to
his analysis) of wayyiqtols in the Bible refer to past events, which accounts for
the majority view that the form has developed from a Semitic prefixed preterite
form. He argues, however, that, because of the problem of induction, confirmatory examples can never confirm a hypothesis, but contradictory ones can
even falsify it (Furuli 2006: 73). Thus, Furuli admits that he allows 6.9% of
the evidence to drive his semantic theory of wayyiqtol! The obvious protest to
this is that Hebrew is an ancient language, attested only in a composite and re51. Janssens (1980: 74) in his review disparagingly states, I have the impression that
Johnsons work is not a progress as compared with the traditional grammar, a good description of which can be found in Joons Grammaire de lHbreu biblique.
129
dacted text that has been vocalized (which is the distinguishing factor between
wayyiqtol and wyiqtol) only hundreds of years after the stabilization of the
text. But this point aside, Sapirs dictum that grammars leak certainly applies
here. Further, Furulis argument that the Masoretes created the wayyiqtol form
and that they made mistakes in writing the form appears prima facie to cancel
out the significance of his 6.9% of counterexamples: these data are simple
errors introduced by the Masoretes; but even so, if the form is simply a
Masoretic invention, how can any of the examples be deemed either erroneous
or representative of the actual language of the texts? In addition, Furuli (2006:
459) admits that his theory is completely at odds with the typological data on
TAM but dismisses those findings, stating that: we should not force modern
views upon a dead language. This comment betrays a lack of understanding
of not only typology but the nature of languages and language universals!
Aside from the unsubstantiated claim that (viewpoint) aspect in Hebrew
differs from the modern linguistic universal and that elements of metaphor
for understanding viewpoint aspect in Hebrew are open to criticism (see Cook
2010), Furulis approach to aspect is fundamentally flawed and contradictory.
First, although he claims that modern views should not be foisted on a dead
language such as ancient Hebrew, he admits that his own analysis is based
on English translations (Furuli 2006: 417). Although by this statement he intends simply to underscore the lack of native-speaker knowledge for a dead
language, it seems all too apparent that his English translations (some quite
wrong) determine his analysis of the Hebrew verbal forms.
Second, his discussion of discourse linguistics is quite illuminating when
immediately followed by his alternative semantic analysis. Having examined
several passages in which the context (adverbial phrases, etc.) affects the aspectual interpretation of the verb form, Furuli (2006: 186) concludes that it
is impossible to know the semantic meaning of most verbs in the Tanakh by
analyzing the clauses and contexts in which they occur. On the following
page (2006: 187), he continues his argument, stating that [O]ur only hope is
to find situations where no other factors than the verb conjugation can cause
a particular characteristic. As an example, he offers his analysis of ( a
wayyiqtol form) in 1Kgs 6:1, which he translates he began to build: The
verb is durative and dynamic, the verb phrase is telic, and the adverbial fixes
the time. But it seems that the small part of the progressive action that is made
visible is caused by the verb form alone, because the only other information
apart from the verb form that is needed is a knowledge of the world (that it
took more than one year to build the temple) (Furuli 2006: 187). In other
words, Furulis analysis of aspect has little to do with the linguistic portrayal
of events; instead, it relates to his preconceived ideas of the character of the
events themselves in the Bible.
130
Chapter 2
2.4.1.4.Summary
The aspectual theories of the past half century all have their roots to some
degree in the Ewald-Driver Standard theory and have all dealt to one degree
or another with issues arising from this theory and the advance of knowledge (both of Semitic and linguistics). One issue is how to explain aspect adequately. Brockelmann (1951: 146) jump-started the discussion by his espousal
of an aspecual analysis using the Latinate terms constative and cursiveterms
that were widely used among European scholars in the decades following his
work. However, B. Johnsons (1979: 30) multiplication of adjectives and metaphors alongside these terms illustrates their limited success.
The other issue is methodogicalnamely, how should a theory of the
BHVS be constructed and how should it deal with the growing body of historical and comparative data. The theories examined above have taken different
tacks with respect to this issue. Rundgren embraced the maturing ideas of
linguistics, drawing on historical and comparative data, but intentionally aiming to describe the BHVS as a synchronic system. By contrast, Meyer made
comparative and historical data paramount, to the point where his theory became untenable because of his attempt to do justice to all of the sundry and (at
points) equivocal data. Michel and similar scholars writing about synchronic
theories, in contrast to Rundgren, chose to ignore completely all historical and
comparative data, describing the BHVS based simply on an inductive study of
the forms. Unfortunately, concomitant with synchronic approaches of this sort
has been the assumption of a one-to-one correspondence between form and
meaning. The result is that the possibility of (partial) homonymy (e.g., yiqtol,
wayyiqtol, and Jussive) among the BH verbal forms is never entertained in
these synchronic studies in the way that it has been considered by diachronic
studies (see, e.g., 2.3.4.2). Instead, in the authors efforts to deal with the
broad range of meanings for a reduced set of verbal conjugationsbecause
of their collapsing of the waw-prefixed and non-waw-prefixed formsthese
synchronic-theory advocates have led to sometimes farfetched explanations of
the verb forms that at once entail concepts from semantics, psychology, and
discourse-pragmatics.
2.4.2. Tense-Prominent Theory
Despite the hegemony of the Ewald-Driver aspectual theory in the 20th
century, the understanding of the Hebrew verbal forms as tense based was not
wholly abandoned. Joon (1923) was one of the notable champions of tense
theory in his grammar (the term future for yiqtol is retained in Muraokas
translation [Joon 2006: 114, 325]). The combination of diachronic interests
with an understanding of the Semitic verb as tensed in Bauer (1910) helped
renew interest in tense-based theories of the Hebrew verb, inspiring the studies
of Blake (1951) and Hughes (1955), which are discussed in 2.4.2.1.
131
Others have dismissed the diachronic approach as being of little value for
informing the way that the BHVS works as a synchronic system. Prominently taking this point of view are scholars who, in a series of studies, have
explained verb choice in BH with reference to syntax, pointing to the obvious
syntactic restriction of the waw-prefixed forms. This line of approach is examined in 2.4.2.2 below.
The rise of relative-tense theory in the mid-20th century (i.e., Reichenbach
1947) inspired a new approach to the BHVS as relative tense. This in turn has
led to more-recent permutations of the relative-tense theory that identify relative tense and modality as the defining parameters of the BHVS. The development of the relative-tense approach is examined in 2.4.2.3 below.
2.4.2.1. Tense and Diachrony
Blake (1944a; 1946; 1951) sought to draw out the implications of Bauers tense theory of Semitic for BH in particular (Bauer 1910; see 2.3.2.2
above). Like Bauer, Blake was dissatisfied with the Ewald-Driver aspectual
theory, stating that [t]he whole [standard aspectual] treatment presents a picture strongly characterized by complexity, obscurity and artificiality (Blake
1951: 1). Therefore, he resurveyed the list of verb meanings found in the
standard treatments in S. R. Driver (1998) and Gesenius (Kautzsch 1910), associating each meaning with either the older or the newer Semitic tense forms.
Although he admitted that the verbs could express certain aspectual nuances,
he claimed that these significations were always subordinate to their primary
tensed meanings (Blake 1951: 2). Blakes inability to surpass the complexity,
obscurity and artificiality of the Ewald-Driver aspectual theory is patently
obvious from his confused concluding paragraph:
The imperfect may denote any tense or mood. . . . The perfect may denote past
tenses but also present or future. . . . Verb forms immediately following [ waw]
have in most cases meanings equivalent to that of the preceding verb. Converted
imperfects and converted perfects may be used independently of any leading
verb. Converted imperfects are regularly past. . . . Perfects with [ w] may have
any of the normal meanings of the imperfect (present-progressive past-futuremodal), but in many cases they are ordinary perfects with past meaning. (Blake
1951: 73)
Hughess study (1955; see also 1962; 1970) was also based on Bauers
theory (1910), but his aim was less confrontational than Blakes: he sought to
illustrate Bauers old and new tense forms in the Hebrew Bible by investigating the syntactic contexts in which the old tense forms might appeartaking
his lead from Bauer (1910: 27). Thus Hughes focused on the verbs that appear
after particles such as , ,
, ,,
and in Genesis2Kings:
It is our thesis that all the Imperfects in past time are vestiges of an old preterite tense of the preformative type (which was found in two forms: yaqtulu
132
Chapter 2
and yaqtul) and are consequently found in stereotyped constructions. They have
been preserved simply because they are in construction with certain particles and
other elements. The preteritive use of the Imperfect is not restricted to instances
with waw consecutive and other particles such as z and rem: additional particles are also used with the Imperfect in a preterite sense. Also, we postulate that
all Perfects in future time are straight (aoristic) future tenses and, like the Imperfects in past time, are to be regarded as archaisms. This futuristic use harks back
to the time when the old afformative verb qatil (qatal) was employed in future
situations. Hence the Perfect is found in future time in stereotypic construction
with other elementsnot simply with waw consecutive but with other particles
as well. It would seem then that the Zeitpunkt does come under consideration in
an evaluation of the usage of the forms and is indicated by factors external to the
verb. (Hughes 1955: 14243)
Aside from a historical interest, neither Blake nor Hughes adds much to our
understanding of tense in BH. Blakes tense system seems as convoluted as
he envisioned the aspectual theory being, and Hughes assumes the validity of
Bauers thesis a priori, so his results really only serve an illustrative end. Both
theories, however, illustrate how difficult tense explanations of the BHVS were
becoming as the level of linguistic sophistication progressed (e.g., cf. Blake
1951 and Brockelmann 1951; or Hughes 1955 and Meyer 1953).
2.4.2.2. Tense and Syntax
Blau in a brief note on the alternation of the waw-prefixed and non-wawprefixed verbal forms signaled a new approach to the question of verb choice
in BH:
To summarize: Biblical prose exhibits a verbal system that denoted tenses, since
the alternation of qal/wayyiqol and yiqol/wqal is due to the syntactic environment (the impossibility/possibility of the use of ww copulative). Accordingly, one will assume a similar system in the spoken language. Deviations in the
usage of verbs in biblical poetry have to be interpreted as intentional archaism.
Since it is impossible to reconstruct such an intricate system as the verbal system
is, from mere archaic features (including, no doubt, pseudo-archaic ones), nothing certain can be inferred from them as to the nature of the Proto-Hebrew verbal
system. (Blau 1971: 26; likewise 2010: 190)
About the same time, Silverman (1973) made similar comments, that qatal
and wayyiqtol are syntactic varieties of past tense, and that yiqtol and wqatal
are likewise both future tense. But he also hinted that this approach could yield
more far-reaching explanations of the BHVS: It is then conceivable that the
more widely studied aspects of completeness (perfect) versus incompleteness
(imperfect) would also be indicated by the placement of the verb within the
clause, and not by its morphology (Silverman 1973: 175).
Revell (1989) has adopted a syntactic explanation for verb choice as suggested by Blau and has attempted to ground it in historical evidence by arguing
133
that, in light of the evidence for a tense system in pre-BH (i.e., preterite yaqtul;
see 2.3) and in post-BH (e.g., Mishnaic Hebrew), [I]t seems likely, a priori,
that the system of the intervening period would also have been one of tense
(1989: 3).52 From this assumption of tense prominence, Revell extends the syntactic explanation not in the direction of aspect, as Silverman had suggested,
but toward modality: modal yiqtol is clause initial, while indicative (tensed)
yiqtol is nonclause initial, and exceptions to this basic pattern represent nonstandard uses of modal yiqtol (Revell 1989: 14). Although Revell was not the
first to observe this syntactic pattern of modality (e.g., Rosn 1969; Niccacci
1987; Voigt 1990), the idea has become more widely disseminated through
Revells students (DeCaen 1995 [see 2.4.2.3]; Shulman 1996; Gentry 1998
[see 2.4.3.2]). With his incorporation of this modal distinction, Revell (1989:
21) is able to add a new dimension to the syntactic explanation: wqatal developed as a syntactic alternative to clause-initial indicative yiqtol, which would
be indistinguishable from modal yiqtol in this position.
Zevit (1988; 1998) has been a proponent of the view that the BHVS primarily indicates tense: qatal is past tense, and yiqtol is present-future tense
(Zevit 1988: 26). At the same time, Zevit takes a different approach to each of
the waw-prefixed forms. With respect to wayyiqtol, he accepts the historicalcomparative argument that it derives from a prefix preterite yaqtul form but
demurs at the notion that examples of preterite yaqtul may be found in the
Hebrew Bible, because the consequent omnitemporal character of yiqtol would
short-circuit the tensed verbal system of BH (Zevit 1988: 30; contra especially Rainey 1986). Instead, he explains yiqtol forms in past contexts as a sort
of historical present use of the form (see Smyth 1956: 422 for this category
in Greek): [I]t [the yiqtol in past context] actualizes a situation by projecting
it into the real time of the speaking voice either for dramatic effect or for emphasis (Zevit 1988: 31). Zevit (1998: 15) explains the significance of wqatal
within the context of his broader treatment of syntax and the qatal form, in
which he proposes that the syntagm waw-subject-qatal following a wayyiqtol
or a qatal clause expresses a past- or present-perfect meaning.
The greatest difficulty with Zevits theory is simply that it is not borne out
by the data (e.g., on waw-subject-qatal, see Gen 4:23). More pertinent to this
discussion, however, are the methodological difficulties in Zevits approach.
First, his position on the use of historical-comparative data is conflicted. While
52.Although Revell does not employ the term relative tense, he nevertheless recognizes
that tense in Hebrew is not absolute: It is important to note that the time reference of the two
categories in relation to the speaker/narrator is not absolute, but is conditioned by the time
reference of the context in which the verb form is used (Revell 1989: 4). I treat Revell in
this section (versus 2.4.2.3), however, because of the importance of his syntactic approach.
His student, DeCaen (1995: 232, 256), follows Revells lead in recognizing tense as relative
(see 2.4.2.3).
134
Chapter 2
135
Word Order
Subject First
Object First
Verb or
Modifiers
First
Subject First
Object First
Verb or Verb
Modifiers
First
Verb First
imperfect (repeated/
distributive)
Conjunctive
Consecutive
and Paratactic
Peckhams theory attempts to take into account innumerable factors affecting the determination of the TAM of the BH verbal forms (e.g., he briefly
discusses the TAM in parallel lines with ellipsis: Peckham 1997: 160), but it
also places too much emphasis on syntax as the determinative feature for TAM
distinctions. As a result, his theory introduces more confusion than clarity,
and it stretches ones imagination that a system such as this could ever have
existed in a living language (see Tropper 1999b). Beneath the clutter of intersecting syntagms, however, Peckhams theory shares features in common with
a broad range of theories of the BHVS, including the tense and syntax theories
discussed above in this section (i.e., tense is determined by word order), the
relative-tense and modality theories discussed in 2.4.2.3 (i.e., qatal and yiqtol
may both express either indicative or modal senses), and even some discoursepromient theories that associate both tense and aspect with the qatal and yiqtol
forms (see 2.4.3.2 below).
2.4.2.3. Relative Tense and Modality
Modern relative-tense theory begins with Reichenbach (1947). One of the
earliest relative-tense models for BH was developed by Barnes (1965) and is
strongly reminiscent of Bulls relative-tense theory (Bull 1960; see 1.2.3.2):
136
Chapter 2
Classical Arabic
Latin
= yaqtulu
= scribo
(non-past, non-anterior)
(Present)
[aspect][tense]
[aspect][tense]
= qatala
= scripseram
(past-anterior)
(Pluperfect)
Figure 2.5. Kuryowiczs schemata of Classical Arabic and Latin verb oppositions (adapted from 1973: 116).
events may be evaluated from multiple points in time, including the actual
present, or positional present 1 in the actual past, or positional present 2
in the actual future (Bull 1965: 131; cf. Bulls schematic, fig. 1.2). Relative
to these temporal evaluation points, qatal and wqatal both denote an event
already fulfilled before ones eyes, while yiqtol and wayyiqtol equally denote
events not in the course of fulfillment (Bull 1965: 7). In like manner to the
tense theories surveyed in 2.4.2.2, Barnes collapses any perceived semantic distinction between the waw-prefixed forms and their non-waw-prefixed
counterparts.
The linguist Kuryowicz (1972; 1973) proposed a model of the Semitic
verbal system that has become influential in relative-tense-and-modality theories of the BHVS and that raises important questions about the relative status
of tense and aspect in the verbal systems of the worlds languages. His basic
argument is that the core of the Semitic verbal systems consists of a privative
opposition between a verb, the primary meaning of which is simultaneous
with the moment of speaking, and another, the primary meaning of which
is action anterior to the moment of speaking (Kuryowicz 1973: 115; cf.
Barness definition of qatal/wqatal and yiqtol/wayyiqtol above).53 Illustrating his point using Arabic as representative of West Semitic (see fig. 2.5),
Kuryowicz argued that [a] binary system like Ar[abic] yaqtulu : qatala excludes not only the category of aspect, but also the category of tense.... The
fundamental relation A [yaqtulu] : [qatala] is neither one of aspect nor one
of tense. Its correct definition is simultaneity (or non-anteriority) versus an53. On privative oppositions, see Crystal 2008: 342, 38687; and above, p.34 n. 28.
137
a. The aspect perfective future, in contrast to the tense futurum exactum (=future perfect), correlates with the so-called prophetic perfect function of BH qatal (Kuryowicz 1972: 87;
1973: 118).
teriority (Kuryowicz 1973: 115). Rhetoric like this was meant to distance
his understanding of the Semitic verbal system from both the Slavic notion of
aspect and the Indo-European concept of tense (1973: 118). 54
In contrast to the range of morphologically expressed TAM values in the
Indo-European languages, Kuryowicz (1973: 116) argued that the West Semitic verbal system expressed the same range of TAM values by contextconditioned functions of the basic binary pair. These functions are illustrated
by the contrasts in fig.2.5, which compares Arabic and Latin. Arabics (and,
by extension, BHs) basic binary opposition between yaqtulu and qatala (:)
expresses relative tense, non-anterior versus anterior; each form of this pair has
secondary functions that create (absolute) tense contrasts with its basic relative-tense functions (: and :), as well as aspectual contrasts between these
secondary functions (:) and between the secondary tense function of yaqtulu
and the primary relative-tense meaning of qatala (:). Hence, Kuryowicz
concludes (1972: 86) that genuine aspect is in Sem[itic] a tertiary function of
the verbal forms. The TAM of each of these secondary and tertiary functions
in West Semitic is illustrated in table 2.9.
54. Kuryowiczs rhetoric has created difficulties in categorizing his theory, illustrated
by Binnicks comment (1991: 435, 438) in one place that Kuryowiczs theory is relative
tense and in another, that he treats the Semitic aspects as absolute time. Kuryowiczs
terminology is a mix of absolute and relative tense: the relation [of Polish] pisze : pisa like
that of il crit : il crivait is an opposition of mere tense (simultaneity with the moment of
speaking : simultaneity with a moment of the past) (i.e., absolute tense; Kuryowicz 1973:
114); and the only non-modal opposition of personal verb forms is yaqtulu : qatala equal to
simultaneity (or non-anteriority): anteriority, tense being context conditioned (i.e., relative
tense) (Kuryowicz 1973: 116). His preference for the labels anteriority and simultaneity adds to the difficulties, because they are also regularly applied to perfect and progressive
viewpoint aspects, respectively (see Bybee, Perkins, and Pagliuca 1994: 54, 13334; Binnick
1991: 28586).
138
Chapter 2
By his claim that tense is always presupposed in a verbal system that expresses
aspect, Kuryowicz raised an important typological question: Which is more
basic in the worlds verbal systems, tense or aspect? This is an especially
55. Kuryowicz also hypothesized that Akkadian may at one time have had morphological distinctions throughout its verbal system, though these oppositions were skewed through
semantic shifts of certain forms, notably the Subjunctive: Present iparras (), Subjunctive
iprusu (), Perfect iptaras (), and Preterit iprus () (Kuryowicz 1973: 11920; cf. Latin in
fig.2.5, p.136). However, Kuryowiczs hypothesis assumes that the Subjunctive form is
basic to the Akkadian verbal system (i.e., a distinct conjugation), a view that is now rejected
by most Assyriologists (see 2.3.2.1).
139
pertinent question because the universal model of tense and aspect underlying
Kuryowiczs Semitic tense theory is the exact reverse of the model proposed
by Bybee and Dahl (1989: 83), as illustrated in fig.2.6.
Three pieces of evidence argue against Kuryowiczs universal tense model
and in favor of Bybee and Dahls (see also Cook 2006a). First, Bybee and
Dahls (1989: 83) model of tense and aspect is based on an extensive crosslinguistic typological study of verbal systems and represents the most common
type of verbal system found in their data. Second, Dahl (1985: 8184) observes
that the relative priority of aspect versus tense in verbal systems is evident
from the fact that tense forms show more morphological similarity with each
other within verbal systems than aspectual forms do. This is illustrated by the
Arabic examples in [2.15] (from Dahl 1985: 83), in which the grammatical
constructions that create a past : non-past opposition are morphologically related, whereas the aspectual distinction is expressed by morphologically distinct forms.
[2.15] Perfective: kataba he wrote
Imperfective: yaktubu he is writing
Past Imperfective: (kana) yaktubu he was writing
Third, based on her hypothesis that the degree of morphophonological
fusion of an affix to a stem correlates with the degree of semantic relevance
of the affix to the stem (relevancy defined as the extent that the meaning
of the category directly affects the lexical content of the verb stem), Bybee
concludes that the category of aspect is most directly and exclusively relevant
to the verb and that its relevancy is reflected in the high degree of fusion between aspectual morphemes and verb stems cross-linguistically (Bybee 1985:
4, 15, 21, 24). Aspect is directly relevant to the temporal character of an event
and alters the meaning of a predication at a more basic level than tense, which
relates to the whole proposition in terms of location in time. Bybees point
is likewise illustrated by the Arabic verb forms in example [2.15], in which
aspectual distinctions are expressed by bound verbal morphology, while tense
is expressed through a periphrastic construction.
The modal component in relative-tense-and-modality theories stems particularly from the contribution of Zuber (1986), which he characterized as
bringing tense and mood/modality into right relation (1986: 29).56 Zuber
argued that the Hebrew verbal system has a binary tense-modality division
between recto-Formen, consisting of qatal and wayyiqtol, and the obliquoFormen yiqtol and wqatal: the recto forms express indicative mood and past
tense (though Zuber stresses mood more than tense); the obliquo forms express
56. Mein Beitrag drft lediglich darin bestehen, Tempus und Modus ins richtige Verhltnis zu bringen.
140
Chapter 2
nonindicative mood and future tense, understanding the meaning of these two
categories as intertwined (Zuber 1986: 27; on the interrelationship of future
tense and modality, Zuber [1986: 15] cites Lyons [1968: 310]). Zuber reluctantly falls back on the medieval waw-hahippuk theory to explain the wawprefixed verb forms, thus interpreting the forms in each group (i.e., recto and
obliquo) as semantically identical and interchangeable (Zuber 1986: 2728).57
Zubers methodology consists in analyzing the forms used to translate the
Hebrew verbs in the Septuagint and Vulgate. He argues that the translators
understanding of the Hebrew verb forms will bring us close to the understanding of the Hebrew verbal system in the Qumran period (Zuber 1986: 38), and
he is persuaded that his statistical examination has sufficiently demonstrated
his thesis (Zuber 1986: 77). However, Zuber never sufficiently addresses other
factors that might contribute to translational agreement between the Septuagint
and Vulgate (e.g., dependency) or the issues involved in analyzing translation
technique between languages with different verbal systems. For example, the
fact that both the Greek and Latin translators employ a Future form to render
yiqtol does not mean that yiqtol is a future tense verb; it merely means that
Future (in such passages) was thought the best equivalent in the Greek and
Latin verbal systems for yiqtol.
Nevertheless, Zuber correctly recognizes that the greatest challenges to his
theory lie in the expression of present tense and past imperfective. In the case
of present tense, the use of both recto and obliquo forms are problematic:
the recto are properly past tense, and therefore their use for present tense is a
makeshift (Ntlosung) use of the forms (Zuber 1986: 89); and Zuber (1986:
89) concedes that the use of the modal-future obliquo forms for indicative present tense presents difficulties for his theory. Similarly, although the translation
of indicative recto forms with Greek and Latin Imperfect (past tense) forms is
perfectly acceptable, the similar rendering of future-modal obliquo forms presents a difficulty for Zubers theory, though he attempts to obviate the problem
by arguing that such instances must be understood as modal oder dubitativ
57. Loprieno (1986) and Rattray (1992) present two notable variations to Zubers model,
both combining this indicative : modal opposition with aspect instead of tense. Loprienos
model features modality (+real : real) intersecting with the privative modal oppositions
from Rundgrens theory (2.4.1.1): unmarked neutral, marked (perfective), and unmarked
negative (imperfective). The +real values are realized by (way)yiqtol (unmarked neutral),
qatal (marked), and yiqtol (unmarked negative), and the real values are expressed by the
Imperative (unmarked neutral), the Subjunctive Cohortative (marked), and the Jussive (unmarked negative; Zuber 1986: 110, 180). Rattrays model is more similar to Zubers, with
the opposition of realis (immediate reality) : irrealis (nonimmediate reality) intersecting a
perfective (or nonprogressive) : imperfective (or progressive) opposition. Thus four categories are realized: realis-perfective qatal and wqatal; realis-imperfective Participle; irrealisperfective Imperative (including Jussive and Cohortative); and irrealis imperfective yiqtol
and wayyiqtol (Rattray 1992: 14950).
141
(Nonindicative) Modal
Nonvolitive (Modal) Volitive (Modal)
yiqtol and wqatal =
future-modal
Imperative,
Jussive, and
Cohortative
and, most significantly, that Iterativ should be understood as properly expressed by the future-modal forms (Zuber 1986: 1012).
One of the most influential scholars in the relative-tense-and-modality camp
is Joosten, who has presented his ideas in numerous articles (1989; 1992;
1997a; 1997b; 1999; 2002; 2005; 2006). The clearest presentation of his theory
appears in Joosten 1997a and 2002, the latter with which I have interacted at
length (Cook 2006a). In his 2002 article, which is directed particularly against
the dominant aspectual theories of the BHVS, Joosten makes his alignment
with Kuryowiczs relative-tense theory particularly evident, claiming with
the latter that a language needs to express tense before it can express aspect
(Joosten 2002: 51). However, already in his 1997a article, his terminology for
the BH verbal forms illustrated his familiarity and favorable disposition toward
Kuryowiczs relative-tense theory: wayyiqtol expresses contemporaneity
with a moment in the past, qatal expresses anteriority . . . with the moment
of speaking, and the Participle expresses contemporaneity with the moment
of speaking (Joosten 1997a: 60). In the latter article, Joosten also acknowledges the influence of Zuber on his thinking (Joosten 1997a: 57 esp.n.21; see
also 1992: 1 n. 1). On these two basesKuryowiczs relative-tense theory and
Zubers modal division of BHJoosten constructs his theory of the BHVS,
illustrated in table 2.10.
One of the foremost challenges to Joostens model of the BHVS is that it
is typologically unparalleled, consisting, as it does of a past (wayyiqtol) :
perfect (qatal) opposition, a perfect (qatal): present (participle) opposition (a
present tense form that is likewise used for past and future expressions), and
two future/modal forms (yiqtol and wqatal) (Cook 2006a: 27). Not only have
no other verbal systems like this been documented in the worlds languages,
viewed from the perspective of diachronic typology, there is no way Joostens
system can be explained in terms of the previous or subsequent configurations
of the Hebrew verbal system (e.g., Joosten [2002: 63] identifies yiqtol as imperfective in proto-Hebrew; elsewhere [1989: 156], he speaks of a historically
earlier use of yiqtol as present tense).
142
Chapter 2
In three other articles (1989; 1992; 1999), Joosten has addressed the same
issues that Zuber recognized as problematic for his tense-and-modality theory.
In the earliest of these articles, Joosten advances the claim that in Biblical
Hebrew the present tense is properly the domain of the predicative participle
(Joosten 1989: 128). He claims that both Subject-Participle and ParticipleSubject word order refer positively to the grammatical present, i.e., they represent an action as contemporary with the moment of speaking (Joosten 1989:
129), in opposition to the widespread view that the participle may appear in
any temporal domain (past, present, or future). Joosten tempers this claim by
noting (1)that the participle may express events that are past or that are still
future as if they are present (i.e., historic present and futurum instans,
respectively); and (2)that the past temporal reference of circumstantial participles (e.g., Gen 19:1) is due to the specific type of subordinate clause (the circumstantial clause), not to the participle (Joosten 1989: 129). The focus of the
remainder of the article is on demonstrating his thesis that Subject-Participle
word order indicates actual present (i.e., an action as actually going on at
the moment of speaking), whereas Participle-Subject word order indicates
factual present (i.e., action represented as fact; Joosten 1989: 130).
Setting aside his word order argument, Joostens association of the predicative Participle with present tense is prima facie problematic in light of the
obvious and plentiful examples of the participle with past and future temporal reference. His attempts to nuance his position with respect to such data
are unsuccessful: given that the grammars agree with him that the temporal
reference of Participles derives from its context, all that Joosten seems to be
claiming is that the Participle defaults for present time. But in this case, the
temporal reference of the predicative participle is still derived from its context,
and his claim that the present tense is in some way the proper domain of the
predicative Participle is at best unhelpful for understanding its role in the BH
verbal system.
In his 1992 article, Joosten argues that the main function of BH wqatal
is the expression of modality (1992: 3), a syntactic alternative (i.e., semantically identical) to yiqtol (1992: 13). Central to his argument are the following points: (1) wqatal should be considered as a separate formal category
with its own function (1992: 7); (2) its use to express iterativity in the past
should be viewed as an extension of its modal function (1992: 7), analogous
with the use of English modal would for past iterative (i.e., habitual). While
I welcome Joostens championing of Zubers view that wqatal expresses
modality, I also have several criticisms. First, he cites Zuber with respect to
wqatals modal functions being to express obligation, potentiality or prediction (Joosten 1992: 3), yet Joosten has never delineated the exact range of
modality expressed by wqatal or how the different meanings relate to each
other diachronically and/or semantically. Relatedly, Joostens use of English
143
would as an analogy is problematic because modal would and past habitual would may in fact be unrelated in English (i.e., homonyms), as some
linguists have argued (e.g., Bybee, Perkins, and Pagliuca 1994: 23839; but
cf. Boneh and Doron 2008; 2010). Finally, I am not persuaded that wqatal
should be treated as a separate conjugation from qatal, either etymologically
or functionally (see 3.3.3.2 for discussion of all three issues).
Finally, in his 1999 and 2002 articles, Joosten particularly addresses the
problem of yiqtol with past temporal reference. In the earlier article, Joosten
surveys typical modal uses of yiqtol with past temporal reference, including
prospective (i.e., future-in-the-past), various nuances of epistemic modality,
and iterative (or habitual) actions. After also examining several problematic
cases, he concludes that, [a]lthough the relatively large number of iterative
instances could be explained as imperfective [cf. Comrie 1976: 25], the absence of clear examples of durative yiqtol (in the past) and the presence of
prospective and modal functions show it preferable to ascribe a basic modal
function to this verb form synchronically (Joosten 1999: 25). This conclusion
is paralleled in his 2002 study, in which he examines possible cases of yiqtols
expressing durative or imperfective action in the present (real present) and
the past (attendant circumstances in the past): The most prominent features
attached to the imperfective in recognized aspect languages are the expression
of real present and of attendant circumstances in the past. Since neither of these
functions is regularly expressed by yiqtol in BH there is no point in classifying
yiqtol as imperfective (Joosten 2002: 53).
Despite Joostens treatments of these problematic datayiqtols expressing
imperfective action in the past or in the presenthis approach is consistently
dismissive: . . . these are modal uses. The seemingly indicative functions of
yiqtol can on closer inspection be explained as being modal as well (Joosten
1997a: 58); the examples that have been invoked to argue that yiqtol does
express the real present can practically all be contested (2002: 54); instances
where yiqtol could be held to express attendant circumstance are infrequent and
generally doubtful (2002: 57). However, the greater problem with Joostens
argument in these two articles (1999 and 2002) is his assumption that if he
debunks the identification of imperfective yiqtol he will likewise defeat the
aspectual theory of the entire verbal system (see Cook 2006a: 27). In fact, such
a result does not necessarily follow a debunking of imperfective yiqtol, nor do
his special pleadings with respect to the difficult data assure that he has successfully disproved the imperfective identification of yiqtol (i.e., he is forced
to admit that imperfective values are at times expressed by yiqtol, though he
wants to see these as marginal uses only).
Several recent dissertations have followed Joostens lead that the BHVS
is modality- and tense-prominent and that syntax has semantic significance.
DeCaens theory (1995) is particularly eclectic, using Rundgrens privative
144
Chapter 2
opposition model to define TAM (see 2.4.1.1), adopting his teacher Revells
tense model (see 2.4.2.2) but expanding it by combining Revells syntactic
observations with Zubers and Joostens binary modal distinction (above, this
section), all within a government-and-binding formalist syntactic framework
(Chomsky 1981).
DeCaens basic schema for TAM is provided in three ternary modelsone
for each parameter: tense = past : non-past (present : subjunctive); aspect=
perfective: imperfective (perfect : progressive);58 and modality = real: irreal
(epistemic : deontic; DeCaen 1995: 205, 210, 218). DeCaen rejects aspect
as an inflectional category in BH: all forms default for perfective aspect, and
imperfective is expressed by the progressive Participle (1995: 22122). He
associates nonindicative modality with verb-first (versus verb-second) word
order: irreal = wqatal; irreal deontic = Imperative, Jussive, and Cohortative;
irreal epistemic = wayyiqtol (1995: 29698). Finally, DeCaen claims that the
BH verbal forms are inflected for tense, as illustrated by his model in fig.2.7,
which he notes is in agreement with the model given by Rundgrens student
Eskhult (1990).
DeCaen offers several important innovations in his study of the verbal system. Most notable is his analysis of word order within a formalist syntactic
framework. Until DeCaens analysis, BH verbal theories, if they took word
order into account at all, described word order in linguistically unsophisticated
ways. DeCaen has brought the proper focus to bear on the order of the subject
and verb and on verb movement within the clause, effected by overt or covert
complementizers (DeCaen 1995: chap.9).
DeCaens other contribution is more problematic than his analysis of word
order: DeCaen posits a second prefix conjugation that includes the Imperative,
Jussive, and Cohortative forms, as well as wayyiqtol, on the basis of morphological and syntactic similarities. The difficulty with this analysis lies in
the mismatched semantics of the deontic Imperative, Jussive, and Cohortative forms and the past narrative wayyiqtol form. DeCaens solution to this
is to posit tense neutralization for wayyiqtol, whereby it is a sequential
form that derives its TAM from the head verb in the sequence (DeCaen
1995: 28492). Unfortunately, this analysis falls under the same censure that
19th-century sequential and inductive theories fall under: (1) wayyiqtol is not
obviously underspecified, as sequential verbs generally are; (2)wayyiqtol
manifestly does not require a head verb, and DeCaen (1995: 286) recognizes that headness is crucial to [his] account of the Hebrew constructions;
58. DeCaen either does not fully appreciate the import of Rundgrens privative opposition model, or he is just careless in his terminology. He correctly labels the unmarked (neutral) category in opposition to perfective aspect as nonperfective, in one instance (1995:
46), but elsewhere he calls it imperfective (e.g., 1995: 32), which results in the incoherent
classification of perfect aspect as a subcategory of imperfective!
145
Inflection (infl)
Suffixed (suff) : Prefixed (pref)
= Past (qatal) = Non-past
Standard Prefixed : Modified Prefixed
Long (pre1)
Short (pre2)
= Present (yiqtol) = Subjunctive (wayyiqtol, Jussive, Cohortative, Imperative)
Figure 2.7. DeCaens ternary model of the BHVS (based on 1995: 105, 282).
(3)he fails to engage the possibility of partial homonymy among the prefix
forms, i.e., morphological and syntactic identity are interpreted as indicating a
single form (cf. Driver [1892] 1998: 77).
At the same time, DeCaens work suffers because of his neglect of semantics, which is illustrated by his comment in his 1999 essay that, the precise
semantic interpretation of this formal model is still an open question (DeCaen
1999: 125). DeCaens lack of clarity with regard to the semantics of wqatal
(i.e., verb-first qatal) seems to be one symptom of his neglect of semantics:
he defines the form as relative past in one part of his discussion (1995: 232,
256), but elsewhere he associates it with irreal modality (1995: 293, 297;
1999: 124).
DeCaen fails to appreciate the semantic interaction among categories of
TAM. For example, he claims that the forms default for perfective aspect, and
that therefore qatal will have a perfective reading except with unbounded
event structures (1995: 256). However, as linguists have pointed out for
languages in general (e.g., C. S. Smith 1999), and I have noted with respect
to BH in particular (Cook 2004b: 253), perfective aspect itself makes events
bounded. Similarly, linguists have argued that perfective aspect is incompatible or canceled when combined with present tense (Bache 1995: 289; Bybee,
Perkins, and Pagliuca 1994: 126), yet DeCaen (1995: 222) claims that the BH
verbal forms default for perfective aspect in every temporal sphere. DeCaens
failure to grapple with the semantic interaction among TAM categories also
undermines his refutation of the aspectual theory of the BHVS: because he envisions tense and aspect as completely independent categories (DeCaen 1995:
181), he finds incompatibilities between aspectual forms and certain temporal
adverbs wholly unexpected (1995: 58). In fact, a decade prior, Dahl (1985:
79) had already identified past temporal reference as a secondary feature of
perfective verbs, invalidating DeCaens conclusion that the aspectual theory
is problematic because there is no pathway of development from perfective
aspect in BH to past tense in RH (DeCaen 1995: 18283).
146
Chapter 2
mood
[+mod]
[mod]
tense
[+vol]
[vol]
[+past]
[past]
aspect
[prog]
Deontic
Epistemic Anterior
Constative
(Juss., Coh., (yiqtol and (qatal and (VS Participle)
and Impv.) wqatal) wayyiqtol)
[+prog]
Cursive
(SV Participle)
147
aspect (Warren 1998: 68), leading to the ordering model for the the BHVS
given in fig.2.8.
Besides this peculiarity, a detailed discussion of Warrens theory is made
unnecessary by its derivational status. His reliance on Joostens ideas (and
Zubers and Kuryowiczs as mediated by Joosten) is particularly evident from
his distinction between volitive (Jussive, Cohortative, and Imperative) and
nonvolitive (yiqtol and wqatal) modality (cf. Joosten 1999: 16); by his constative: cursive distinction, signaled by subject-Participle word order (cf. Joosten
1989); and by his choice of labels, such as anterior for qatal and contemporaneous for the Participle (cf. Joosten 1997a: 60). Although DeCaens theory
is not completely consonant with Joostens, Warren (1998: 78) notes that it
is highly susceptible to a modal interpretation of yiql (cf. DeCaen 1995).
However, he is somewhat uneasy about his acceptance of DeCaens explanation of the waw-prefixed verbs as continuation forms: Our adoption of
DeCaens argument for why the continuation forms should cross-match with
the main-clause forms remains problematic. DeCaen does not show clearly
how he has moved from mood-neutralization to the ascription of new modal
values to wqal and wayyiql (Warren 1998: 94). In any case, the value of
Warrens theory does not lie in his understanding of the TAM of the BHVS
but in his empirical investigation of the range of modality and its expression
in BH, which may be drawn upon without partaking of the weaknesses in his
adoption of the ideas of Joosten, Kuryowicz, and DeCaen.
Penner (2006) focuses on the Qumran verbal system, using as a foil the
claims of most standard textbooks that BH primarily grammaticalizes aspect and that MH primarily grammaticalizes tense (Penner 2006: 211).
Based on his synchronic, empirical, statistical analysis of the verbal forms in
a number of sectarian writings from Qumran, Penner concludes that, though
there is a strong correlation between the verb forms and both absolute tense
and modality (recto and obliquo), Qumran Hebrew is mainly tense-prominent:
Under this hypothesis, past events and present states are expressed using qatal
or a waw-prefixed verb following a qatal, and future events and present actions
are expressed using yiqtol or a waw-prefixed verb following a yiqtol (Penner
2006: 213).
Although Penner concludes that absolute tense and modality are prominent
in Qumran Hebrew, he is nevertheless quite dependent on the relative-tenseand-modality theories for many of his ideas: his use of Zubers modal categories of recto (qatal and qatal+wyqtl) and obliquo (yiqtol and yiqtol+wqtl)
(Penner 2006: 17071);59 and his classification of future as properly modality, using arguments similar to Joostens against an aspectual view of yiqtol (Penner 2006: 177). At the same time, he draws on more recent theories,
59. Penner notes (2006: 21) that Zubers conclusions may well be more applicable to
Qumran Hebrew than BH (see Zuber 1986: 38).
148
Chapter 2
notably Hatavs (1997; see 2.4.3.2 below) for ideas concerning modality and
sequentiality (e.g., Penner 2006: 9697, 2225).
Besides the fact that Penners theory partakes of the weaknesses of Joostens
and the other relative-tense-and-modality theories, as described above, his empirical method is suspect when compared with Furulis contemporary study
(based on a broader database that includes Qumran Hebrew; see 2.4.1.3),
which arrives at very different conclusions: both Furuli and Penner attempt an
empirical, statistical analysis of the verb forms in context (Furuli 2006: 186
87; Penner 2006: 1012) and arrive at contradictory results, Furuli positing
some unique form of aspect for Hebrew (with just two main conjugations
prefix and suffix), and Penner proposing tense-prominence in Qumran Hebrew
(Furuli 2006: 46264; Penner 2006: 21213). It appears prima facie that both
Furuli and Penner have found what they were looking for; that is, their interpretation of the data was guided by what they expected a priori to find, which
accounts for their very divergent results based on overlapping data sets.
2.4.2.4.Summary
The modern revival of tense theory of the BHVS against the background
of the hegemonic Ewald-Driver aspectual theory has been widely influential
because of two significant insights in particular. First, modern tense theory
has brought to bear an ever-more-sophisticated analysis of the syntax of the
BHVS, theorizing that certain TAM distinctions may in fact be syntactically
rather than morphologically marked. Second, modern tense theory has been
instrumental in bringing modality into the TAM discussions of the BHVS. One
of the most important insights, not yet fully explored in the literature, is the
modal status of wqatal.
At the same time, however, modern tense theories suffer from several
shared weaknesses. First, an absolute tense theory appears prima facie to be
an impossible model for the BHVS, because the forms clearly function in all
three temporal spheres, even if there are certain notable patterns in their distribution. Second, modern tense theories have largely protested the diachronic
theories of the BHVS, and like the synchronic aspectual theories, thereby miss
the possibilities of homonymy in the BHVS (see 2.4.1.4). The results are
semantically implausible explanations for explaining the broad range of meanings for particular forms (e.g., DeCaens tense-neutralization of modal yiqtol
to explain past-narrative wayyiqtol). Third, modern tense theory is most problematic when viewed from a typological perspective. This is recognized by
tense theorists themselves, who have all struggled to some extent to explain
the use of modal yiqtol as past and present imperfective. Generally, however,
the data have not been sufficiently explained, the typological data have been
selectively used (e.g., see Cook 2006a: 2628 on Joostens use of typological
data), and statistical analyses have been used to dismiss the relevant data as
149
negligible (e.g., see Cook 2006a: 2829 on Joostens use of statistics; see comments above, p.148, on Penners and Furulis inductive, statistical approach).
However, the faulty efforts of modern tense theories serve as important
guides to a sound analysis of the BHVS. First, the BHVS must be analyzed as
a system rather than in terms of individual verb forms in order to avoid proposing arbitrary and typologically unparalleled TAM models for BH, such as
are found in tense theories (e.g., Joosten, DeCaen) or in the assumption that
understanding the system rests solely in explaining yiqtol (e.g., Joosten 2002
and my response, Cook 2006a). Second, statistical analysis is not as crucial to
the argument as sound typological analysis.60 Inductively and systematically
examining the verb forms according to context has been unsuccessful for both
aspectual and tense theories (e.g., Furuli and Penner): this approach seems to
discover what the scholar a priori assumes is there, whether tense, aspect,
modality, or none of these. Instead, a theory of the BHVS should be judged
by whether it presents a typologically credible model of the verbal system
in light of the abundance of data on verbal systems in the worlds languages.
With reference back to Kuryowiczs theory, which pointedly raised the typological question of the relative priority of tense and aspect in the verbal
systemsthe tense theories examined here have already been shown to be
typologically less credible than an aspectual model of the BHVS.
2.4.3. Discourse-Prominent Theory
Discourse analysis entered linguistic terminology with Harriss 1952 articles by this title (1952a; 1952b).61 He used the term in a fairly restricted sense
to refer to the analysis of discourse through breaking it up into its fundamental
elements. Now, however, the term is without a doubt one of the most widely
used and loosely defined terms in the entire field of linguistics (Prince 1988:
164). Nevertheless, most approaches to discourse analysis, and particularly
those interacted with below, share two general characteristics. First, they are
unified by a common object of inquirya discourse/text as the verbal representation of a communicative act (Brown and Yule 1983: 6). This characterization entails two common assumptions of discourse analysis: linguistic
analysis should not be restricted by the traditional grammatical boundary of the
sentence but should extend beyond this to the discourse level; discourse is a
social (communicative) act, not simply an artifact.62 Second, the general goal
60. The data must, of course, be interacted with. Andrasons recent spate of articles
(2010; 2011a; 2011b; 2011c; forthcoming) falters on his obvious lack of acquaintance /
discomfort with the Hebrew text, aside from the theoretical deficiencies of his semantic
foundation (compare with my approach in 3.1).
61. Text-linguistics is the more commonly employed term among European linguists.
62. Although discourse analysis (as linguistics generally) has tended to give normative
status to the study of spoken discourse, analyses of written discourse are becoming more
150
Chapter 2
151
Imperfective
simultaneity or chronological overlapping
descriptive, non-action situations
background information
low transitivity
nonpunctual/durative; view of a situation
or happening not necessarily complete
subject identity consistent, presupposed frequent change of subject
human topics
nonhuman topics
realis
irrealis
affirmative
negative
and correlated it with temporal succession (i.e., foreground is temporally successive; background is not). In turn, Hopper (1979: 214) viewed temporal succession as effected by perfective aspect, leading to a three-stage correlation:
foreground is temporally successive and is marked by perfective aspect; background is not temporally successive and is indicated by imperfective aspect
(see also Hopper 1982).63 In another influential article, Hopper and Thompson
correlated foregrounding with transitivity: The grammatical and semantic
prominence of Transitivity is shown to derive from its characteristic discourse
function: high Transitivity is correlated with foregrounding, and low Transitivity with backgrounding (Hopper and Thompson 1980: 251). These correlations among discourse functions and verb forms, illustrated in table 2.11, form
an important foundation for the discourse analysis of the BHVS and explain
the predominant focus on foreground-background and temporal succession in
these theories.
Third, and a logical outcome of the previous two characteristics, discourse
analyses of biblical literature have emphasized the importance of describing
verbal forms in terms of the discourse-pragmatic functions. For a segment of
the field, this approach derives from Weinrichs discourse theory of European
languages (2001: 43; 1st ed. 1964), in which he argues that syntax should be
described in terms of providing a preliminary sorting (Vorsortierung) of the
63. I should note that Hoppers understanding of temporal succession and his simplistic (in retrospect) association of it with the perfective-imperfective aspectual distinction is
merely indicative of the time period in which he wrote, during which investigations into
temporal succession and the semantic components that contribute to its expression were still
nascent (see discussion in 1.5). For an up-to-date discussion of temporal succession and its
role in BH, see Cook 2004b; and 4.2 below, and the references cited there.
152
Chapter 2
discourse world for the speaker and listener. This preliminary sorting is defined
by the distinctions between speech (Besprechen) and narrative (Erzhlen),
between between background and foreground (Relief), and among backward, neutral, or forward Perspektivthat is, temporal reference (Weinrich
2001: chaps.24).
Longacre, whos ideas have been influential especially among Bible translators, shares the same basic perspective as Weinrich: I posit here that (a)every
language has a system of discourse types (e.g., narrative, predictive, hortatory,
procedural, expository, and others); (b)each discourse type has its own characteristic constellation of verb forms that figure in that type; (c)the uses of given
tense/aspect/mood form are most surely and concretely described in relation to
a given discourse type (Longacre 2003: 57). By most surely and concretely,
Longacre means to set his discourse theory in contrast to traditional semantic
theory; but as Talstra (1997: 86) states, the debate over whether to describe
the verbal forms at the discourse-pragmatic level or semantic level is not
one of principle, but one of priority. Although Talstra argues for the priority
of a discourse-pragmatic analysis and that the categories of tense, mood
and aspect operate only within the limits of the textual organization as it is
created by text-grammatical categories, he also states that it is important to
remain open to the possibility of relating text-level and clause-level categories
to each other within one system of grammatical analysis (Talstra 1997: 85
86). In Niccaccis theory, TAM values are correlated with specific discourse
functions: We can affirm that verb forms have fixed temporal reference when
they are verbal sentences and/or indicate the mainline of communication both
in narrative and in direct speech. On the other hand, they have a relative temporal reference when they are nominal clauses and indicate a subsidiary line
of communication (Niccacci 1994b: 129).
All of these approaches, however, are subject to the criticism of an unhealthy circularity because of their notable lack of any semantic foundation:
correlations between discourse functions and specific verbal forms presume
the independent determination of the semantic value of the verbal forms; otherwise, how can we know what is being correlated with a given discourse function? This inherent circularity when discourse analysis is given priority over
semantic analysis has been noted with respect to both Weinrichs discourse
theory (Bache 1985: 22) and Longacres discourse studies of BH (Hatav 1997:
21; see also Heimerdinger 1999: 98) and is discussed in detail in 4.1 below.
Although there are a variety of approaches to discourse analysis of biblical
literature, there are two particularly influential schools: the Longacre school
and the Schneider school.64 Longacres discourse analysis, which has been
widely used in the analysis of previously unstudied languages and for Bible
64. Not surprisingly, the Longacre school has predominated among American scholars,
while the Schneider school has been more influential in European scholarship.
153
+Contingent
Succession
Contingent
Succession
+Agent-Orientation
Agent-Orientation
NARRATIVE
Prophecy
Story
PROCEDURAL
How-to-do-it
How-it-was-done
+Projection
BEHAVIORAL
Hortatory
Promissory
Eulogy
EXPOSITORY
Budget proposal
Futuristic essay
Scientific paper
+Projection
Projection
Projection
translation, stems from Pikes tagmemic theory (Pike 1967; Waterhouse 1974).
The approach is empirical and functionalthat is, language is part of human
behavior (Dawson 1994: 7576). The central concepts of tagmemics (not restricted to its application to discourse) are slot and filler and the hierarchical
character of language. Slot and filler or class and set refer to the paradigmatic
(or associative) and syntagmatic relationships in language (see de Saussure
1959: 12227), referred to in tagmemic theory as tagmemes and syntagms:
tagmemes combine at various levels of language structure to create syntagms.
Thus, Longacre (2003: 302) explains that the two create a correlative relationship: tagmemes are expounded by their role in syntagms, and syntagms are
composed of various tagmemes. This part-whole relationship occurs at each
level of language structure, usually by employing syntagms of a lower level
as tagmemes at the next level up, forming a hierarchy of morpheme, word,
phrase, clause, sentence, paragraph, and discourse (Longacre 2003: 302; Dawson 1994: 82).
Longacres tagmemic-based discourse analysis begins with the assumptions that identifiable linguistic patterns exist at the discourse level and can
be described, and that text-type is the stongest determining factor at the
discourse (macro-syntactic) level of tagmemes lower in the hierarchy (i.e.,
micro-syntactic level; Dawson 1994: 23; see Longacre 2003: 57, quoted
above this section). A difference of some importance for the present study is
that Dawson, in the exposition of his Longacre-based theory, states that correlating micro-syntactic tagmemes with macro-syntactic roles in no way
lessens the micro-syntactic identities (Dawson 1994: 23). By contrast, Longacre, quoted above, argues that verb forms are most surely and concretely
described in terms of their role in a particular text-type (Longacre 2003: 57).
This uncertainty about the status of micro-syntactic versus macro-syntactic
constructions is exacerbated within discourse studies of BH because of the
existing challenges involved in describing the verbal system.
154
Chapter 2
155
Narrative
Speech
wayyiqtol
yiqtol/Imperative
x-qatal
x-yiqtol
backward forward
x-qatal
qatal
wqatal
backward zero
forward
156
Chapter 2
157
Niccacci, in contrast to Schneider, also accords a greater place to semanticsthe neglect of which Talstra criticized Schneiders theory (Talstra 1978:
174; see Niccacci 1990: 2122). Unfortunately, Niccaccis attempt to combine discourse-pragmatics and semantics has been characterized as muddled
categories and mixing paradigms (Exter Blokland 1995: 23), because it is
unclear in Niccaccis theory how Schneiders discourse model of the Hebrew
verb (see table2.14, p.155) and his own semantic distinctions, which derive
from Bartelmus (1982; see Niccacci 1990: 165 and 2056), are related. Not
only this, but it is unclear which takes precedence in defining the BHVS, and
even Niccaccis description of the semantic categories is unclear. In one place
he states, [T]he function of the verb form or a grammatical construction is
dependent on its morphology and meaning. Context, style and literary composition have also to be taken into account since they combine to determine the
meaning of a text (Niccacci 1990: 163). Two pages later, he refers back to
this statement, explaining:
We have already stated that semantics is of importance, even if only secondary,
in determining the function of a verb form or grammatical construction. Further
clarification is now required. In fact, in the account given so far the following
criteria based on semantics and interpretation have emerged which play a part in
the choice of verb forms or constructions in texts: simultaneous or prior action,
single or repeated action, emphasis (mode of action, Aktionsart). (Niccacci
1990: 165)
158
Chapter 2
159
160
Chapter 2
it can be called upon to shore up his insufficient discourse-pragmatic explanations of verb choice. Thus, when faced with explaining several examples of
wqatal as terminal paragraph markers, in place of the more usual qatal with
this function, he explains that the narratological reason why these clauses
were used instead of the more usual QAAL terminal clause is related to their
inherent semantic and aspectual meanings, since discourse pragmatics and the
semantic fields of the verbs work together to provide the sense of the clauses
(Heller 2004: 439: n. 17). Given this sort of statement, it is a wonder that only
in such instances does Heller find it necessary to say anything about the semantics of the verbal forms, and even then his statements are of the vague and
uncertain character of this quotation (e.g., Heller 2004: 311 n. 44, 324 n. 48).
Del Barcos recent discourse study (2003) has connections with both Long
acre and Niccacci: following Longacre, del Barco posits notional discourse
types, such as narrative and predictive; at the same time, his taxonomy for
qatal and yiqtol is organized on the basis of word ordera particular focus
of Niccaccis work. The primary distinctive of del Barcos work is that he
has chosen for his database the preexilic minor prophets (Hosea, Amos, Micah, Nahum, Habakkuk, and Zephaniah). As a result, his emphases differ from
prose-narrative-based discourse analyses. This is seen in his more extensive
treatment of the wqatal form than other studies vis--vis his treatment of
wayyiqtol (Barco del Barco 2003: chap. 3) and in his identification of discourse types that are particular to prophetic speech: predictive, exhortative,
descriptive (or expository, lamentive, discursive narrative, and discursive
interrogative-rhetorical (2003: 23641).
In line with previous discourse theories, del Barco distinguishes wqatal
and wayyiqtol from the non-waw-prefixed forms primarily in terms of sequentiality of events. The primary function of wqatal is the expression of
sequential events in predictive discourse, in which it has a future temporal reference. However, the form may appear as well in predictive discourse without
a sequential sense, often following a lead yiqtol form (Barco del Barco 2003:
89). Outside predictive discourse, wqatal may appear in descriptive types of
discourse (e.g., expository, lamentive) without an explicit temporal reference
(2003: 97). Del Barcos (2003: 126) analysis of wayyiqtol is also similar to
previous studies: when the form appears in prophetic discourse, it is usually
headed by qatal and expresses past tense, sequential narrative events, usually
embedded in speech.
Del Barco discovers the same sorts of variation in function/meaning for the
non-waw-prefixed forms as may be found in previous studies. He emphasizes
that interpreting the function/meaning of qatal and yiqtol depends on paying
attention both to the discourse context and to their position within the clause
(Barco del Barco 2003: 158). Qatal appears independently (i.e., not headed by
waw or subordinating conjunction) with an anterior or perfect meaning, rarely
161
Tense-Aspect
qatal
yiqtol
wayyiqtol
wqatal
impv-juss-coh/yiqtol
w+impv-juss-coh/wqatal
appears with waw (which would be indistinguishable from wqatal), and may
appear with a variety of functions/meanings following subordinate conjunctions (2003: 158). Yiqtol has three different functions/meanings clause initially: a modal form (i.e., jussive), the antecedent of a chain of wqatal forms,
and at the head of either clause in a protasis-apodosis construction or parallel
bicola (2003: 19495). Yiqtol prefixed with waw shows a close functional/
meaning connection with the preceding clause, often with a continued jussive
sense. Finally, yiqtol functions variously following subordinate conjunctions,
similarly to the case of qatal (2003: 195).
Given the similarity of treatment in del Barco with the other discourse
theories surveyed here, it is open to the same or similar criticisms (see Cook
2004a). I offer an example from his work to illustrate the difficulties inherent
in his approach. Del Barco identifies the occurrence of wqatal and yiqtol in
the midst of a past-tense narrative sequence in Amos 4:78 as an atypical case
(un caso atipico), because the forms generally belong to predictive discourse
where they have a future temporal reference. Hence, he reluctantly concludes
that these forms express future tense in a predictive discourse that intrudes in
the midst of a past-tense narrative discourse (Barco del Barco 2003: 8889).
However, the traditional semantic approach to the BHVS has no such problem
with the occurrence of these forms in narrative discourse, since it recognizes
as part of their semantic makeup the expression of past habitual (or customary) action (Waltke and OConnor 1990: 53334). As with other discourse
analyses, del Barco runs into problems because the BH verbal forms do not
align with discourse functions as uniformly as he expects. When deviations
from their expected functions are encountered, discourse analysts have no
recourse to semantics except in vague and ad hoc ways (see on Heller, above),
and if a semantic explanation is rejected, they can deduce, as does del Barco,
conclusions that are directly counter to what traditional grammars have long
recognized.
In the final decade of last century, there was a confluence of opinions on
the Hebrew verb that resulted in a flurry of publications, all seeking to refine a basic shared model of the BHVS, illustrated in table 2.15. This basic
model consists of two or three parameters by which the BH verbal forms are
162
Chapter 2
distinguished: a tense-aspect semantic distinction, a discourse-pragmatic distinction, and in some models a modal semantic distinction between indicative
and nonindicative. One of the earliest models of this sort is by Gropp (1991),
which quite closely resembles the generic model in table 2.15 (p. 161). He defines the tense-aspect distinction as relative tense (anterior) and distinguishes
between the waw-prefixed forms and non-waw-prefixed forms (including volitives: indirect versus direct) in terms of sequential, which he understands in
terms of contingent temporal succession, following Longacre (Gropp 1991:
50; see Longacre 1996: 89); and he includes a third, modal distinction (volitive) to differentiate the indicative and nonindicative forms, as defined in the
traditional grammars (for direct and indirect volitives, see Joon 2006:
114, 116).
Besides the criticism of distinguishing forms in terms of discoursepragmatic function (sequential), Gropp creates difficulties by defining both
qatal and wayyiqtol as [+anterior] (relative tense), because wayyiqtol almost
always implies anteriority to the moment of speaking (i.e., absolute tense;
1991: 55). His solution is ad hoc: In order to account for the relationship
between the perfect [qatal] and the narrative [wayyiqtol] we need to posit a
semantic rule such that +anterior in the context of +sequence, is to be interpreted as +past, or in other terminology, the interaction between +relative past
and +sequence converts the form semantically to an +absolute past (Gropp
1991: 55). Similarly, Gropps distinction between direct and indirect volitives
in terms of sequential is less than perfect, because the semantic range of indirect volitives is arguably greater than just contingent temporal succession (see
Waltke and OConnor 1990: 57576), not to mention that Muraoka (1997a)
has called into question the validity of this volitive distinction.
Buths model (1992) is an example of the two-parameter variation of the basic model in table 2.15. In contrast to Gropps form-meaning correspondences,
Buth understands these parameters to be functional categories. This is evident
from his preliminary remark that, Like any human language, Hebrew is able
to make time and aspect distinctions. Reports to the contrary can be ignored
(Buth 1992: 96). His functional approach is further evidenced in the broadness
with which he defines these parameters. He characterizes the tense-aspect parameter in terms of verb forms for definite events (that is, past or perfective
or decisive or contrary to fact) versus verb forms for indefinite events (future
or imperfective or potential or repetitive) (Buth 1992: 103). Buth refers to the
discourse-pragmatic parameter as thematic continuity, which encompasses
both temporal succession and foregrounding: the waw-prefixed forms signal
thematic continuity, whereas the non-waw-prefixed forms signal discontinuity (1992: 1014). Unfortunately, the broad, functional basis of Buths model
eschews the very questions we are attempting to answernamely, the semantic notional categories marked by the BH verbal forms. Buth addresses neither
163
Endos understanding of aspect, as reflected in this quotation, is quite different from the understanding found in the great majority of linguistic works.
Although aware of Comries advocacy for the terms perfective: imperfective
instead of complete: incomplete (Comrie 1976: 18), Endo (1996: 320) rejects
this suggestion but then proceeds to confuse complete: incomplete with the
significantly different opposition of completed: incompleted; this is evident
from the fact that he treats perfective: imperfective and complete: incomplete as distinct categories (e.g., Endo 1996: 55; cf. Comrie 1976: 1819, who
equates them) and at the same time views complete : incomplete as functionally indistinguishable from past: non-past (cf. Comrie 1976: 18, whose
164
Chapter 2
discussion implies that the completed: incompleted pair may easily be confused or equated with the past: non-past temporal distinction in just the sort of
way that Endo confuses tense and aspect categories).
Second, Endos description of the parameter of sequentiality is impene
trable. Although he does not offer a definition, he clearly intends to distinguish sequentiality as a syntactic relation at the discourse level (i.e., between
clauses) from the more customary understanding of sequentiality in biblical
discourse linguistics as temporal/logical succession of events: the parameter
of sequentiality and non-sequentiality is purely syntactical, relating to the flow
of the story as discourse function: the non-sequential form stops the flow of
the story, whereas the sequential form lets the story flow on (Endo 1996:
70). Endo argues that it is the verb form itself in the waw-prefixed forms that
has this syntactic marking of sequentiality, rather than being effected by the
conjugation, based on the diachronic argument, going back to Bauer (1910),
that Hebrew had two sets of suffix and prefix conjugations, distinguished by
stress placement:
Biblical Hebrew may be of the kind [of language] where sequentiality in the
verb form is clearly distingiushd [sic] from the function of the conjugations.
That is, it seems to have particularly sequential verb forms, since, ... it is highly
probably that biblical Hebrew has two sets of conjugations in each temporalaspectual distinction: QATAL and (waY)YIQTOL for past (complete); YIQTOL
and (w)QATAL for non-past (non-complete). The latter conjugations usually
appear in the sequential context. (Endo 1996: 68)
Neither does word order have any semantic significance; rather, Endo classifies all word-order variation as related to the pragmatic feature of topicalization (Endo 1996: 320). Finally, Endo also rejects the widespread view
(generally concomitant with the temporally successive understanding of sequentiality) that verb forms distinguish anything like foreground and background in discourse (Endo 1996: 322). The uncertainties regarding Endos
syntactic understanding of sequentiality underscore the confusion that abounds
regarding this parameter (see further below).
Gentrys model (1998) of the BHVS represents a culmination of the generic semantic and discourse-pragmatic theory in table 2.15 (p. 161). His
model, given in table 2.16, includes four parameters in total: two semantic
parameters (modal and tense-aspect), and two discourse parameters (sequentiality, and narrative versus direct speech). In addition, he incorporates the
patterns of negation that interact with word order and restrictions on modal
forms (Gentry 1998: 157).
Gentrys understanding of the BHVS combines insights from several quarters. First, he accepts the diachronic argument that yiqtol and (way)yiqtol derive
from two separate prefix conjugations in West Semitic (Gentry 1998: 1013;
165
wayyiqtol
wqatal
Projective Modality
Perfective
Imperfective
Jussive
( + Jussive)
yiqtol
+ yiqtol)
(
waw + Jussive
waw + yiqtol/
(wqatal)
Narrative
Conversation (direct speech)
see 2.3.4.2 above); he argues that both diachronic and synchronic analysis of
the verbal system are necessary (Gentry 1998: 9). Second, he argues that the
verb forms morphologically indicate an aspectual opposition of perfective:
imperfective, while tense is indicated by the combination of verb form and
discourse context; for example, in direct speech, the perfective normally
expresses past tense, and the imperfective forms non-past tense (Gentry 1998:
1420). Third, Gentry closely follows Buth in his understanding of sequentiality as a discourse-pragmatic parameter. Thus, the sequential verb forms are
employed to encode continuity or to foreground information in the discourse
(Gentry 1998: 13). Fourth, Gentry follows his teacher Revell and the work
of other Revell students (notably Shulman 1996) in his treatment of modality (assertive/indicative versus projective): only the Imperative morphologically marks projective modality; otherwise, the modal opposition is signaled
by word orderprojective modality being marked by initial position in the
clause (Gentry 1998: 2124).
Gentry helpfully interacts with many of the previous theories of the BHVS,
making it quite easy to determine the relationship of his model to others, both
previous and contemporary. For example, he credits Hendel (1996) with his
definitions of aspect and modality and cites Gropps definition of tense (Gentry 1998: 14, 21). As already mentioned, he derives his understanding of sequentiality from Buth (1992) and his syntactic treatment of modality from
Shulmans study of volitives (1996). In addition, Gentry distances his model
from previous theories of the BHVS, noting that the tense-aspect parameter
is unconnected with the sequence of verb forms (Gentry 1998: 20; compare
19th-century waw-inductive theories, above, 2.2.1). He also departs from the
consensus view that the -suffix on the Cohorative is a marker of projective
modality, preferring instead to identify the Cohorative with the on other
166
Chapter 2
forms (notably the 99 instances of wayyiqtol with final )as a single morpheme
related to the Akkadian Ventive -a (Gentry 1998: 2530; see above, 2.3.4.2).
At the same time, Gentrys attempt at comprehensiveness results in an
unhelpful category confusion similar to the confusion that permeates many
models of this variety (i.e., those akin to the basic model in table 2.15 above):
the model mixes together parameters defined by semantic meaning with parameters of discourse-pragmatic functions and, in some cases, syntactic distinctions, without adequately distinguishing these linguistic levels. This sort of
confusion seems to lie at the root of protestations such as the humorous quip
by Rainey (1986: 7) that [t]he ancient Israelite farmer certainly knew when
to milk his cow and his language was adequate to explain the routine to his
son, to which an apt reply would be Buths remark, cited above: Like any
human language, Hebrew is able to make time and aspect distinctions. Reports
to the contrary can be ignored (Buth 1992: 96). Semantics (meaning) must be
distinguished from syntax and, above all, discourse pragmatics (function) in
the analysis of the BHVS.
A major distinctive feature of discourse-prominent theories of the BHVS
is sequentiality, which has been defined (or not defined) in various ways, notwithstanding the general consensus that it involves the notions of temporal
succession and foreground. However, there has been a good deal of confusion over sequentiality, evident in works such as Endos (see above) and Lis
(1999), due to a lack of rigorous definition (e.g., Endo) or due to confusion
between similar phenomena related to sequentiality (see Cook 2004b: 248).
Lis definitions illustrate the potential for confusion:
It may be useful to bring three definitions together. An iconic clause is one that
occurs in the temporal and/or logical order of the events represented in the narrative. A sequential/consecutive clause is an iconic clause that is marked for
sequence. A narrative clause is the least marked iconic syntagm in the narrative,
and normally also the most frequently occurring iconic syntagm. (Li 1999: 4,
italics added)
167
and narrative clause. Several approaches to the issue of sequentiality are illustrated here.
Dempster (1985) attempts to avoid just the sort of confusion that Li seems
to encounter by distinguishing between syntactic sequentiality (sentence sequencing) and semantic sequentiality (predicate sequencing): sentence sequencing refers to the wa-clause connecting, whereby a clause prefixed with
wa- presupposes a previous clause with which it is in sequence (Dempster
1985: 40); predicate sequencing presupposes sentence sequencing, but further
indicates that the predication in the clauses so connected present a coherent
sequence of past events (Dempster 1985: 33, 51). Dempster further identifies
the sequencing constraint as interactive with word order variation: The normal pattern in which the prefixed verb occurs is used to present the majority
of verbal predications. If it is removed narrative ceases to exist. Two formal
conditions regulate the use of the prefixed verb in this pattern: sentence position (first) and sentence sequence (wa). This verb is restricted to linguistic
contexts in which it is the first grammatical constituent of syndetic sentences
(Dempster 1985: 6869). Thus, speaking of wayyiqtol versus qatal, Dempster
states that the verbs themselves in the different patterns do not signal any
significant semantic difference ... but the pattern itself is the primary signal of
the semantic consecution (Dempster 1985: 68, 96). In other words, the syntax
is the marker of both sentence sequencing and predicate sequencing.
Dempsters discourse study focuses on four supra-sentential constraints,
derived from Gleasons class lectures on discourse analysis: sequencing, theme,
reference, and lexical cohesion (Dempster 1985: 38 n. 1). DeRouchie (2007)
has adapted these four constraints in his discourse study of Deuteronomy 511,
though otherwise, DeRouchies study is quite different from Dempsters in that
DeRouchie primarily follows Longacres discourse theory and incorporates a
semantic component based on Cook (2001; 2002; 2004b) in an effort to avoid
the pitfalls of simple form-function correlations (see DeRouchie 2007: 2528).
With regard to Dempsters sequencing constraint, DeRouchie explains:
What the present study terms text logic and foregrounding, Dempster used
the term sequencing and distinguished between sentence sequencing (i.e., connection) and predicate sequencing (i.e., semantic succession). But as has been
recently argued by J. A. Cook [Cook 2004b], not only is sequencing to be distinguished from temporal succession, but also temporal succession is most appropriately seen as a context-determined feature and not a grammaticalized one.
(DeRouchie 2007: 78 n. 62)
Several questions and problems are raised by these statements. First, DeRouchies wording implies that he is equating his text logic category with
Dempsters sentence sequencing, and foregrounding with Dempsters predicate sequencing category. However, as stated above, Dempsters predicate
sequencing is the equivalent of temporal succession, which DeRouchie seems
168
Chapter 2
wqatal
+
yiqtol
qatal
qotel
(Participle)
Modality
Progression
Perfect
169
system, expressing progressive aspect, the defining feature of which is inclusionthat is, the ability to include other events in its reference time (Hatav
1997: 89); finally, (4) wayyiqtol is defined by sequentiality and the absence of
Hatavs other feature.
Wayyiqtol (and to a lesser extent wqatal) and sequentiality are the central concerns of Hatavs theory. Interacting with semantic theories regarding
the movement of reference time (see 1.5 above), Hatav (1997: 6) defines
the waw-prefixed forms as always presenting bounded events and therefore
advancing the reference time in discourse. Hatavs relating of temporal succession (sequentiality) to boundedness is no doubt correct (see Hatav 1989;
Depraetere 1995); however, her strict association of boundedness in BH with
the waw-prefixed forms and her opposing claim that qatal never expresses
boundedness denies both the complexity and sometimes the indeterminacy of
boundedness (see C.S. Smith 1999) as well as the obvious counterexamples in
the biblical text, both of wayyiqtol without boundedness and of qatal expressing boundedness (Cook 2004b: 25167). Hatav has mollified these claims in a
more recent article (2004), allowing that qatal may express bounded events by
borrowing its reference time from another clause in the context, having its
reference time determined by an adverbial phrase, or taking the speech time as
a default reference time (Hatav 2004: 51318). Thus, Hatav begins to sound
more and more similar to writers who make a narrative: speech distinction
between wayyiqtol and qatal (cf. Gropps statements on the reference time of
these two forms, above; and Goldfajn and Kawashima, below this section).
However, a problem that persists in Hatavs recent compositional treatment
of wayyiqtol and sequentiality (2004) is her identification of the underlying
prefix form in wayyiqtol as modal yiqtol, when the historical and comparative evidence points to separate origins for these two conjugations (see 2.3
above). She argues that wayyiqtols semantics arises from the restricting of its
modal meaning by two prefixed morphemes: the doubling ay morpheme acts
similarly to the definite article in a noun phrase, anchoring the verbal event
in the familiar actual world (Hatav 2004: 5007); the wa morpheme effects
sequentiality by creating a new reference time (Hatav 2004: 51012). This
analysis is simply a more linguistically sophisticated and semantically oriented
revision of DeCaens earlier analysis of the form (1995), on which, see my
critique above (2.4.2.3).
Alongside sequentiality, the other major trait of discourse-prominent theories is their distinction between narrative and (direct) speech, although this
distinction may be more or less covert in some theories (e.g., Gropps distinction between wayyiqtol and qatal may be correlated with the reference-time
distinction between narrative and direct speech, above), or it may be obscured
by subcategories within each (e.g., Heller, above). Goldfajns theory (1998),
notably, makes the narrative-speech distinction centrally important: narrative
170
Chapter 2
67. Goldfajn (1998: 100) also find significance in word order changes that accompany
verb alternation as signaling temporal shifts, for example, the subject-verb order with qatal
as interrupting the narrative flow of verb-subject wayyiqtol (see 2.4.2.2. above).
171
2.4.3.3.Conclusion
Discourse-prominent theories as a whole have been among the most influential force in the discussion of the BHVS. In this concluding section, I want
to underscore some distinctive features of these theories and provide a preliminary critique of discourse-prominent theory vis--vis semantic theory, which
will provide a starting point for the much more extensive discussion in chap.4.
The central features in discourse-prominent analyses of the BHVS are the
oppositions of sequentiality: nonsequentiality, foreground: background, and
narrative: speech. Unfortunately, these phenomena are not consistently definedespecially sequentialityif they are defined at all, and they are often
assumed to be coterminus with each other; that is, sequential events are foregrounded events in narrative, whereas direct speech features nonsequential and
often background events (versus the surrounding narrative material). The first
portion of chap.4 (4.12) is therefore devoted to examining these phenomena
and their interrelationship.
Despite a large degree of uniformity among discourse-prominent theories,
different conclusions emerge depending on whether the model takes as its
starting point the Longacre approach or the Schneider. The most notable difference is that the Schneider school focuses on the narrative-speech distinction,
whereas the Longacre school works with a variety of text-types. The latter is
open to criticism with respect to both the utility (see Niccacci 1994b: 119) and
the soundness of the determination of these text-types. It is especially unclear
how such notional text-types are to be distinguished from literary genre (so
Michael OConnor, private communication, 2006). The result of these different starting points is subtle but important: in the Schneider school, the major
divide in the BHVS comes between wayyiqtol and yiqtol (or in some cases qatal), the former belonging to narrative as the latter does to speech; by contrast,
in the Longacre school, the primary distinction made is between the temporally
successive waw-prefixed forms and the other forms. What appears evident here
is that Longacres approach places Ewalds 19th-century consecutive theory
on a new foundation of discourse analysis. This assessment is confirmed
by Longacres semantic understanding of the waw-prefixed forms (e.g., he
begins with GKC and faults it not for being erroneous as much as for being
incomplete; Longacre 1994: 50; 2003: 62); and studies such as Grosss (1976),
in which he argues for a consecutive understanding of wayyiqtol in a more
sophisticated linguistic framework than Ewald.
My central critique of discourse-prominent theory with regard to the BHVS,
which has arisen at various points throughout the preceding survey, is that
the functional identifications of the verb forms apart from attention to semantics is methodologically problematic. In the most extreme cases, a semantic
component of certain verbal forms is denied altogether (e.g., Baayen 1997 on
qatal). More often, however, the theories proceed from discourse-pragmatic
172
Chapter 2
173
chronic data. The verbal forms that lie at the center of this debate are wayyiqtol
and yiqtol, because of the great importance often attached to the diachronic
evidence for a short *yaqtul and a long *yaqtulu underlying these two forms,
respectively.
Unfortunately, Hebraists, just as much as biblical scholars, have yet to resolve the question how to apply a synchronic approach to a composite (i.e.,
temporally disparate) text such as the Hebrew Bible. More pointedly, however,
the perspective that lies at the center of this debate is problematic. In earlier
studies, I objected to the terms of this debatethat diachrony and synchrony
are mutually exclusive options for the analysis of the BHVSon the basis of
criticism of de Saussures strict distinction between diachrony and synchrony,
and I pointed out that Hebraists have failed to remain consistently synchronic
even when they have consciously sought to take this sort of approach (see
Cook 2001: 122; 2002: 19194). Since then, I have come to recognize that
the problem with the synchrony-diachrony debate lies at a more basic level of
linguistic understanding: Hebraists have generally incorrectly assumed that
wayyiqtol must be construed as related both formally and semantically to yiqtol if diachronic data is made inadmissible. However, synchronic studies
(cf. Michel 1960; Kustr 1972) as well as diachronic demonstrate that these
forms differ in meaning and function to such a degree as to make them distinct
though partially homonymous (disregarding for the moment the waC- prefix)
conjugations regardless of their etymology. An argument for homonymy does
not need to rest on any diachronic data but can to be supported by a synchronic
analysis of the distribution of the two forms.
The second issue, closely correlated to the first, is meaning versus function
in the description of the BHVS. Concomitant with the shift from diachronic to
synchronic description, there has been a movement away from description of
meaning (often based on etymology) to the function of verb forms (within a
synchronic system). This reflects to some degree the development in linguistics generally (from a concern for semantics to pragmatics) and is represented
by the move toward discourse analysis, which is concerned more with the
function of the verb forms within a discourse than the meaning attached to the
individual form itself.
Objections to the abandonment of meaning for verb forms in favor of functional analyses have been raised by linguists (e.g., Comrie 1985) and Hebraists (e.g., Joon [and Muraoka] 2006: xviiixix) alike on the grounds that
verbs have a discernible meaning apart from their discourse context. Although
context crucially informs the meaning/function of words, to deny thereby that
a word has any discernible meaning or a fuzzy meaning at best goes against
the well-established Fregean principle of compositionality that meaning is a
product of the contribution of individual components. What is the contribution
of the verbal morphology to the interpretation of the form in contextwhether
174
Chapter 2
175
more or less successful theoretical lens. How then are competing theories to
be judged? As with all interpretation, the questions that they must answer are:
Does this theory explain more data more satisfactorily than any other? But to
whose satisfaction? Generally, each interpreters theory is to his/her own satisfaction the best theory, so there needs to be some objective means by which
competing models may be judged. In chap.3, I argue that typology appears to
be invaluable for evaluating competing models, because it asks in light of the
ever-increasing data from the worlds languages whether a particular theory is
typologically believable/reasonable or anomalous.
Typology, therefore, appears to be an especially useful tool for the study of
ancient languages, because it provides an external means of judging the results
of inductive studynamely: Is the model of the language typologically believable or is it typologically anomalous? The value of typology has already been
shown preliminarily in my criticism of relative-tense theories above: they are
typologically anomalous (2.4.2.3). A typological approach suited particularly
to the chronologically disparate writings of the Hebrew Bible is presented
inchap.3.
Chapter 3
176
177
We do not take the structuralist position that each language represents a tidy system in which units are defined by the oppositions they enter into and the object of
study is the internal system the units are supposed to create. Rather, we consider
it more profitable to view languages as composed of substanceboth semantic
substance and phonetic substance. Structure or system, the traditional focus of
linguistic inquiry, is the product of, rather than the creator of, substance.
178
Chapter 3
and diachrony can be understood in total isolation from each other. But this too
is incorrect. A particular synchronic stage of a language is, in part, a product of
childrens reanalyses of an earlier synchronic stage. In that way, synchrony and
diachrony are inseparable. On reconsideration of the matter, I reject my earlier endorsement of Heine, Claudi, and Hnnemeyers panchronic rhetoric
(Cook 2001: 122; 2002: 19194) and adopt the argument that, unless language
systems change through some means other than cross-generational and crosslectal transmission, the notion of panchronic is nonsense (Joseph and Janda
1988: 194).4
All this said, it is not surprising that typology has long embraced a diachronic dimension of analysis (see Greenberg 1978), given its empirical methods and association with functionalisms substantialist approach to language.
Bybee, Perkins, and Pagliuca (1994: 34) list four reasons for their taking a
diachronic approach to typology of the TAM categories, three of which I list
here by way of endorsement. First, diachrony increases the explanatory power
of the theory. This is affirmed by the more recent comment on causal explanations in typology by Moravcsik (2007: 39): [T]here is no need to choose
between synchronic and diachronic accounts: synchrony is what diachrony
explains. Second, given that language is constantly changing, it is unnecessarily limiting to examine only a thin synchronic slice (Bybee, Perkins, and
Pagliuca 1994: 4). Rather, and third, [S]imilarities among languages are more
easily seen from a diachronic perspective. I would add to these arguments that
the diversity of the data for the BHVS (i.e., cross-generational as well as crosslectal) argue in favor of attention to diachronics. Arguably, in the absence of
native speaker linguistic tests, diachronics (and particularly diachronic typology) remains the only truly viable external control on the analysis of BH
grammar.5
Finally, Saussure distinguished between signal (signifiant) and signification (signif) unified in the idea of the sign: I call the combination of a
concept and a sound-image a sign. . . . I propose to retain the word sign [signe]
to designate the whole and to replace concept and sound-image respectively
4. Andrasons recent (2010; 2011a; 2011b; 2011c; forthcoming) advocacy for panchronic explanations of the BHVS only underscore the weakness of such an approach, which
pays little attention to the BH verbal system as a system and is content with a taxonomy of
unranked functions for individual forms.
5. I have made this case in Cook 2012, where I explore the possibility that the use of
diachronic typology as an external control may enable us to date portions of the Hebrew
Bible relatively based on the development of the verbal system, among other things. This
approach using diachronic typology as the means of validation is directly in contrast to the
statistics-based approach followed in recent years, aided especially by computer databases
(see chap.2).
179
180
Chapter 3
classical model.9 Unfortunately, it has too often been characterized as the only
real alternative to the classical model (for a critique, see Wierzbicka 1990;
1996: 14869) and as the inevitable choice of theory for functionalists:
It might be expected, then, that functionalist accounts of language would reject
the classical, Aristotelian concept of classification, under which entities can be
classified in terms of sets of necessary and sufficient features, in favour of some
version of the approach to non-discreteness which, prompted by Wittgensteins
philosophical observations on family resemblance models of meaning, have
been developed, largely by psychologists, into what is now generally known as
prototype theory. (Butler 2003: 28)10
181
182
Chapter 3
183
184
Chapter 3
fallen out of use. As a result, the possibility that the text has been altered
(by scribes or the Masoretes) through grammatical misunderstanding (not to
mention inadvertent changes) must always remain open. (This seems to be especially true in Psalms, where the temporal reference of many portions is ambiguous, and therefore the context is less helpful in deciding the cases.) Cases
such as these are especially problematic when we are dealing with unvocalized
texts, such as the Dead Sea Scrolls, where an analysis of the prefix forms may
presume a priori in which contexts each form is grammatically acceptablea
seemingly problematic assumption (Penner 2006).
A similar sort of a priori understanding of the BHVS seems to underlie the
widespread and favored use of statistics (e.g., Furuli 2006; Penner 2006). The
use of statistics provides a misleading sense of objectivity and definitiveness
to the enterprise, because it mimics empirical science. However, in the end
the statistics only serve as a tally of the interpreters subjective and often pre
determined semantic interpretation of the forms of the BHVS. Statistics cannot
serve to validate semantic interpretation, which still partakes of human enterprise. Statistics are only valid when they tally objectively measurable things,
such as the number of times a distinctly Jussive form of the verb appears
clause initially (e.g., Shulman 1996). Statistical studies are even unnecessary
for gathering the requisite data for analysis, because their results are readily
available in the standard reference grammars, old and more-recent, including
examples of both standard meanings (e.g., Waltke and OConnor 1990; van
der Merwe, Kroeze, and Naud 1999) and inexplicable uses (e.g., Gesenius
1910; Joon 1923). The real challenge, then, is validation: How can one validate a theory of the BHVS?
One answer to this question is to avoid it and settle instead for a taxonomy
of syntactic and discourse uses of the verb forms. The approach of syntactic
taxonomy as theory can be traced back to Moran (2003) and seen in the works
of Rainey (e.g., 2003a) and his students (e.g., Zewi 1999). Not only do these
taxonomies fail to provide a theory of linguistic structure, they lead to unwarranted implications that the correlation between form and syntactic context
is causal (i.e., this specific meaning can be explained as contingent on this
particular syntactic context). This is the case because they ultimately can only
offer a list of specific meanings for forms when what is desired is a general
meaning for each form to which the specific meanings can be reasonably related.14 The same criticism may be leveled at discourse studies, which merely
14. Despite his employing a much more sophisticated semantic theory, a similar criticism can be leveled at Andrason (2010; 2011a; 2011b; 2011c; forthcoming), who is content
with a taxonomy of meanings for each BH verb form that is cognitively coherent without
any concern for either relating the varied functions to a general meaning or ranking the
meanings in any way for a given form. Thus, he essentially advocates the abandonment of
any sort of general-specific distinction in meaning (e.g., Jakobson) in favor of an amalgam
185
186
Chapter 3
Bickel (2007) notes that in recent years typology has escaped its narrow
confines of being merely another tool for determining universals of grammar
and shifted its focus from the former question of whats possible? to whats
where why?15 Because of the long-standing conviction that typological distributions are historically grown, this shift of inquiry has brought with it a
concomitant increased interest in historical explanations and what is referred
to as diachronic typology (so Croft 2003: chap.8). Croft (2003: 233) explains:
In diachronic typology, synchronic language states are reanalyzed as stages
in the process of language change. Diachronic typology therefore seeks not
simply to classify language types but shifts between types, because language
states are seen as the product of type transitions and diachronic processes in
general (Bickel 2007: 239). Thus, as Moravcsik (2007: 39) states, [T]here is
no need to choose between synchronic and diachronic accounts: synchrony is
what diachrony explains.16
A field closely associated with diachronic typology (it emerged from the
latter according to Croft 2003: 253) is grammaticalization. The term grammaticalization refers to both a type of diachronic change in languagewherein
lexical items become grammatical items or grammatical items become more
grammaticaland a framework for examining diachronic changes (Hopper
and Traugott 2003: 12). The process of grammaticalization has been recognized for almost a century; the term was coined by Antoine Meillet in 1912 (Fr.
grammaticalisation; see Hopper and Traugott 2003: chap.2). However, the
more recent development of grammaticalization theory, composed of various principles recognized with respect to grammaticalization processes (see
Hopper 1991; Bybee, Perkins, and Pagliuca 1994: 922), has increasingly been
severely criticized (see esp.Newmeyer 1998: chap.5). Campbell (2001: 113)
has noted that grammaticalization has only a derivative status, because it can
be reduced in every instance to a combination of diachronic processes known
independently from grammaticalization theory (e.g., reanalysis and analogy).
For this reason, grammaticalization theory has, at best, a heuristic value
within diachronic typological research (Campbell 2001: 158). Similarly, Newmeyer (1998: chap.5) has argued that grammaticalization is epiphenomenal
and not properly a theory. More importantly, he criticizes the circularity
of constructing a theory of grammaticalization based on reconstructions that
presume the validity of the theory. My use of grammaticalization studies in
15. This shift in inquiry is complete enough to prompt Nichols (2007: 232) to write:
Typological theory is almost entirely unconcerned from distinguishing possible from impossible languages.
16. To illustrate her case, Moravcsik (2007: 37) cites Dryer (2007: 246), who notes that,
although language function often describes typological explanations for language structure,
this is simply a loose way of saying that language function reflects diachrony, which in turn
explains language structure (see there for his examples).
187
188
Chapter 3
Pre-Latin
*?
*kanta bh umos>
Latin
French
cantabimus
cantare habemus> chanterons allons chanter> ?
wolde wanted>
OldMiddle English
wolde wanted
wolde (auxiliary)>
Present-Day English
would (auxiliary)
process or renewal (Croft 2003: 253).18 This latter process can be readily illustrated in the tendency of languages to develop new periphrastic expressions
that may exist for a time alongside older synthetic forms, as in the case of the
Latinate futures, the development of which is schematized in fig.3.1.
Another way in which new constructions may develop is through divergence, in which a form develops a new grammatical meaning without immediately discarding the older function for the form: When a form undergoes
grammaticalization from a lexical to a grammatical function, so long as it is
grammatically viable some traces of its original lexical meanings tend to adhere to it, and details of its lexical history may be reflected in constraints on its
grammatical distribution (Hopper 1991:22). This development is especially
well illustrated in the case of the grammaticalization of English wolde/would,
schematized in fig.3.2.
In the intermediate stage of this change, we find the following rare example,
[3.1], of both the past inflected meaning of the form and its auxiliary function
in a single passage.
[3.1] a Darius geseah t he overwunnen beon wolde, a
when Darius saw that he overcome be would, then
wolde he hiene selfne on m gefeohte forspillan.
wanted he him self in that battle kill:inf
When Darius saw that he would be overcome, he wanted to
commit suicide in that battle. (Hopper and Traugott 2003: 48)
18. Moreno Cabrera (1998: 224) argues that the principle of unidirectionality is better
described as irreversibility.
189
specific
known
specific
unknown
irrealis
nonspecific
question
indirect
negation
direct
negation
conditional
comparative
free choice
Figure 3.3. Conceptual space for indefinite pronouns (Croft 2003: 135).
valaspecific
known
specific
unknown
irrealis
nonspecific
question
conditional
indirect
negation
comparative
akr-, br-
semdirect
negation
free choice
190
Chapter 3
191
192
Chapter 3
The relative rarity of this construction (i.e., plus infinitive) with stative
roots20 may be explained by the freedom with which stative roots express an
inchoative based on the logic of the discourse context, as illustrated in example
[3.3].21
[3.3] Neh 13:19
Then, when the gates of Jerusalem became shaded/darkened
prior to the sabbath, I said . . ..
The coda phasal aspects, which express the discontinuation of a situation, include the cessative (with [telic] events) and completive (with [+telic]
events). Both of these are marked lexically in BH with the same sort of infinitival construction as the onset phasal aspects: cessative aspect is expressed by
or ( cease), and completive aspect uses ( finish), as illustrated
in examples [3.4ab].22
[3.4] a. Gen 18:11
Abraham and Sarah (were) old, advanced in days; the
manner of women had ceased to be for Sarah.
b. 1Kgs 9:1a
Then, when Solomon finished building the house of Yhwh
and the house of the king . . ..
Phasal aspects that apply to the nucleus of the event structure affect the
progress of a situation, through either repetition (iterative and habitual) or extension with or without a pause (resumptive and continuative). The expression
of these aspects in BH is less uniform than in the cases of the other phasal
20. I have found only two stative and one passive examples: with the Infinitive of
(Gen 10:8; 1Chr 1:10) and with a Niphal Infinitive of ( be stopped up). Gen 9:10 features a (stative) verbless clause as the complement of .
21. Other inchoative examples are: Gen 26:13 ( become great); Deut 32:15 (
become fat); Judg 1:30, 33, 35, etc. ( become); 1Sam 4:1 ( become slack); Ezek
31:5 ( = become high; become many; become long); Neh 13:19 (
become overshadowed).
22. Examples with : Gen 11:8; 18:11; 41:49; Exod 23:5; Num 9:13; Deut 23:23;
1Sam 12:23; 23:13; 1Kgs 15:21; Isa 1:16; Jer 44:18; 51:30; Ps 36:4; Ruth 1:18; 2Chr 16:5.
Examples with : Jer 31:36; Ezek 34:10; Hos 7:4 (with Participle instead of Infinitive).
Examples with : Gen 17:22; 18:33; 24:15, 19, 22, 45; 27:30; 43:2; 49:33; Exod 5:14;
31:18; 34:33; Lev 16:20; 19:9; 23:22; 26:44; Num 7:1; 16:31; Deut 7:22; 20:9; 26:12; 31:24;
32:45; Josh 8:24; 19:49, 51; Judg 3:18; 15:17; 1Sam 10:13; 13:10; 18:1; 24:17; 2Sam 6:18;
13:36, 39; 1Kgs 1:41; 3:1; 7:40; 8:54; 9:1; 2Kgs 10:25; Is 10:18; Jer 5:3; 26:8; 43:1; 51:63;
Ezek 43:23; Amos 7:2; Ruth 3:3; Dan 9:24; 12:7; 1Chr 16:2; 2Chr 4:11; 7:1; 29:29.
193
He sent them away from Isaac, his son, while he (was) still
alive.
Iterative and habitual are semantically similar but refer to the application of
patterned repetition to a minimal event type and any event type, respectively.
Although I have not found a clear iterative example in the Hebrew Bible, habitual expressions are numerous and may be expressed in a variety of ways. In
most cases, the logic of the way the situation fits within the discourse context
leads to a habitual interpretation of yiqtol or wqatal), as in example [3.6] (see
additional discussion in chap.4).24
[3.6] Gen 29:2b3
. . . because from that well they would water the flocks. Now
the stone was large that was upon the mouth of the well, and
all the flocks would gather there and they would roll the stone
from upon the mouth of the well and they would water the
flock and they would return the stone that was upon the mouth
of the well to its place.
In other instances, temporal adverbial expressions disambiguate and/or reinforce the habitual interpretation of the yiqtol or wqatal forms, as in examples
[3.7ab].25
23. One could argue that again represents yet another type of phasal aspect, such as
repetitive, since there is no agreed-upon list of possible phasal aspects (see chap.1 n.27).
However, given the specific definition of phasal aspect in 1.7.3.2, again does not qualify,
and it seems preferable to retain the stricter definition rather than allowing every adverbial
modification of a situation to become a type of phasal aspect.
24. Other habitual expressions are: with wqatal, Exod 33:7; 1Sam 7:1516; 2Sam
17:17; 2Kgs 3:4; Amos 4:78 (on the latter, see comments in Cook 2004a: 337); with yiqtol,
Gen 2:6; 31:39; Exod 33:7; 40:36; 2Sam 15:32; Num 9:18.
25. See also 1Sam 1:7; 2:19; 7:16; 1Kgs 5:25.
194
Chapter 3
This is what Job would do all the time.
b. 1Sam 1:3
Now that man would go up from his city periodically to
worship and to sacrifice to Yhwh Sabaoth in Shiloh.
The confusing issue that arises with regard to a habitual expression is deciding whether it is a matter of phasal aspect or irrealis or nonindicative mood.
Although logically, it makes some sense to consider habituals acting on predicates to be phasal aspect, this classification is also motivated by the origin of
the study of phasal aspects in Slavic languages (see 1.3.2). The absence of a
system of phasal affix markers in BH and recent arguments connecting habituality to irrealis mood make a modal analysis of habituality in BH preferable
to a phasal aspectual analysis. Below (3.3.3), I analyze habitual yiqtol and
wqatal as irrealis mood.
A full treatment of phasal aspect is beyond the scope of this book and less
to the point than an examination of the semantics of verbal marking itself (i.e.,
verbal morphology and syntax versus lexical expression of TAM). The brief
preceding overview illustrates well how the expression of phasal aspect fits
with other parts of the BHVS, especially with respect to its interaction with
situation aspect, to which I turn next.
3.2.2. Situation Aspect: The Stative-Dynamic Opposition
Situation aspect classifies situations in terms of their internal temporal constituency. In chap.1, I discuss the development of this category from Aris
totles comments on types of action and from some of the classifications of
situation aspect that have been developed (1.3.1). In constructing my model
of TAM, I have adopted Rothsteins minimal classification of four situation
types distinguished by two features ([stages], [telic]; Rothstein 2004;
1.7.3.1). However, the practice of classifying BH situations in relation to
this model is much more difficult than the distinction of situation types in the
abstract. In addition to the difficulties already endemic to classifying situation
types (see 1.3.1 for discussion), the lack of native speakers hampers the use of
linguistic acceptability tests to distinguish among types. This is especially the
case in distinguishing accomplishments ([+stages, +telic]) from achievements
([stages, +telic]).26
26. For example, are the verbs
( he will take . . . in order to place) in 1Sam
8:11 accomplishments (the plural locations, with his chariots and with his horsemen, imply
[+stages]) or achievements (the verbs to take and to place are generally [stages])?
195
Most relevant to the theory I am constructing is the morphological distinction in BH between (as it has traditionally been understood) stative and
dynamic (i.e., activity, accomplishment, achievement) situations. In brief, in
the Qal binyan (which is the only binyan in which the distinction is observable) in the suffix-pattern, qatal-conjugation dynamic verbs have an a theme
vowel (e.g.,
he kept) while stative verbs have an *i(>e) or *u(>o)
theme vowel (e.g., he is old and he is small); correspondingly, in the
prefix-pattern conjugations (yiqtol, wayyiqtol, Jussive-Imperative) dynamic
verbs have a *u(>o) theme vowel (e.g.,
he will keep/is keeping;
keep!) while stative verbs have an a theme vowel (e.g., he is becoming/will be old;
he is becoming/will be small; see G.R. Driver 1936:
chap.7 for a full listing of stative roots).27 Several semantic characteristics
accompany this morphological pattern, reinforcing the identification of the
pattern as a dynamic-stative distinction. First, stative verbs generally lack an
active Participle form but instead may be encoded either verbally (i.e., verbal
inflection), or adjectivally (the ms form is identical with the Qal 3ms qatal).
Second, the stative verbs show a distinct pattern of interaction with the qatal
and wayyiqtol conjugations: conjugated in qatal, stative verbs default to a present stative meaning, whereas conjugated in wayyiqtol, they always express past
states. (Context may dictate an inchoative meaning for statives in either qatal
or wayyiqtol.) All these issues have significance for understanding the development and meaning of the BHVS and are discussed in turn.
By contrast to my above characterization of these morphological and semantic patterns as having a dynamic (or fientive or active) : stative distinction
(so Joon 2006: 41, 112), other labels for this distinction have been suggested (both for Semitic in general and BH in particular), including transitive:
intransitive (so Gesenius 1910: 11819; see W.Wright 1962: 1.28 regarding
Arabic) and active voice : passive/middle voice (Tropper 1998a: 182; Joosten
1998: 207). The reason for so many competing explanations is the interaction
of voice, transitivity (and valency), and situation aspect and the patterns of
correlation among such parameters and the vowel pattern in BH (and Semitic).
To clarify the situation adequately (though not completely), it is helpful to
27. The dynamic-stative vowel pattern is regularly obscured in verbs with a pharyngeal
or largyngeal consonant in the second or third positions of the verbal root, which tend to
effect an a theme vowel, regardless of the underlying vowel pattern (e.g., be strong;
cf. in pause, 2Sam 1:23;
he stretches/will stretch out x), and in verbs with final
glide consonants that undergo contraction (e.g., *rabiya > rh, it was numerous; cf.
*galaya > glh, he uncovered). This pattern is part of a larger observable pattern, referred to as the Barth-Ginsburg law (see Joon 2006: 41 n. 12), involving vowels of both
the suffix-pattern and the prefix-pattern conjugations: dynamic verbs have a-a suffix vowel
patterns and a-u prefix vowel patterns; stative verbs have a-i/u suffix vowel patterns and
i-a prefix vowel patterns (in the Masoretic tradition, the attenuation of a-vowel in a closed,
unstressed syllable levels the distinction in the prefix vowel).
196
Chapter 3
have in place some basic definitions for dynamic and stative situation aspect,
transitivity and intransitivity, and active, middle, and passive voice. These are
given in example [3.8].28
[3.8] a. The dynamic : stative opposition classifies verbs/predicates
according to their situation aspect: either they express a
state or some variety of dynamic event, such as activity, accomplishment, or achievement. In addition to states being
used dynamically, such as when they are combined with
certain phasal aspects, such as inchoative (e.g., He was old
> He became old), certain verbs are naturally stative and
dynamic, such as smell: He smelled (the roses).
b. The transitive : intransitive opposition categorizes verbs/
predicates according to the number of inner participants
that they require: intransitive (subject), transitive (subject
and object), ditransitive (subject, object1, and object2).
Stative verbs are mostly intransitive, but there are regular
exceptions, such as love (BH )and hate (BH )as in
Jack loves/hates Jill.
c. The active : passive opposition classifies verbs/predicates according to the role of the subjectwhether agent
or patient. The subject of an active predicate is the agent,
whereas the subject of a corresponding passive predicate
is the patient, as in The wind struck the house versus The
house was struck (by the wind). The middle has elements of
both the active and passive voices in that the grammatical
subject is both agent and patient.
The interaction among these parameters is evident in Klaimans (1991) study
of grammatical voice. However, a comparison between the vowel-pattern
marked distinction in BH and the systems that Klaiman examines (ergativeabsolutive, active/dynamic-stative, and active-middle) makes it evident that
BH distinguishes dynamic and stative. On the one hand, it is readily apparent
that the class of verbs with i/u theme vowel in BH is neither consistently passive nor consistently intransitive, and Joostens argument (1998) that this class
of verbs is middle voice is based on the view that middle voice is essentially
intransitive (see Klaiman 1991: 45). While the majority of the i/u class of verbs
28. Transitivity is a narrower category than valency in that as it deals entirely with the
inner participants familiar from traditional grammar: direct objects and indirect objects.
While a verb may only license one of each type of inner participant (multiple noun phrases
will be construed as compounds), it may require (semantically and/or grammatically) any
number of free participants (e.g., locative prepositional phrase with certain verbs of motion); valency treats both types of participant.
197
198
Chapter 3
the qatal conjugation was restricted to past temporal reference even with stative roots (Segal 1927: 150; Kutscher 1982: 131).31
Another typological argument tied to the dynamic-stative analysis not only
serves to confirm the dynamic-stative analysis of the theme-vowel alternation
presented here but contributes to the analysis of the verbal conjugations (particularly, qatal and wayyiqtol; see 3.2.3.1 and 3.4.1). The argument is based
on the observable pattern of interaction between stative verbs and the qatal
and wayyiqtol conjugations: stative verbs default to present temporal reference
when conjugated with qatal; that is, absent contextual indicators otherwise,
stative verbs are interpreted as present states.32 By contrast, stative verbs conjugated in wayyiqtol consistently show a past temporal reference. This contrast
is substantiated by a statistical analysis of 49 stative verbs: 33 conjugated in
31. Segal (1927: 150) states that in the Mishnah qatal with stative roots is restricted to
expressing past states, and the conjugation no longer exhibits any future temporal reference.
Kutschers statement (1982: 131) that the perfect now [i.e., in RH] denotes only past action
is consonant with Segals earlier assessment. Nevertheless, Prez Fernndez (1997: 108)
objects that M. H. Segal overstates his claim that forms like can never have a present
significance in R[abbinic] H[ebrew] [i.e., I know vs. I knew], for in fact, we find in rabbinic
literature certain idiomatic turns of phrase, such as
, in which the present is clearly
signfied. However, none of Prez Fernndezs examples (see 1997: 11617) involve stative
roots. Ridzewski (1992:160) offers an example of a stative root in qatal, which he categorizes as Prsens:
. However, he translates it with a past inchoative sense (wir
luden Schuld auf uns), which is consistent with the above assessment that qatal is restricted
to past temporal reference even with stative roots. While these examples, therefore, do not
contradict the claim that qatal is restricted to past temporal reference in Rabbinic Hebrew,
they do illustrate the complexity of the verbal semantics.
32. Joon (2006: 112a) characterizes the present temporal reference of statives in qatal
as their primary meaning.
33. The selection is based on the lists in Joon (2006: 41) and G.R. Driver (1936:
4647). An attested stative meaning was the primary parameter for selection; thus some
very common stative roots were included despite an attested active Participle (these are designated by ptc. in the list); those with verbal adjective are marked with adj. and the predominant dynamic meanings in some forms is glossed as a development (>) of the original stative
meaning: ( ptc.) love; ( adj. ) be high; / be strong > prevail; ( adj.
)be great; /
( adj. )
cling(?); ( adj. ) be low; ( adj.) be fat;
(adj.) be old; ( adj.) be strong > prevail; be leavened; ( adj.) be polluted;
( adj.) be lacking; ( adj.) (be) delight(ed) in/with; be ashamed; ( adj.
) be terrified > tremble; be clean; ( adj.) be good; ( adj.) be unclean;
( adj.) be dry; ( adj.) be afraid; ( adj.) be able; ( adj.) be weary; ( adj.)
be afraid; ( adj.) be asleep > fall asleep; ( ptc.) be clothed > clothe; ( adj.)
be heavy; be advantageous, be proper; ( adj.) be full > fill; be foolish;
be pleasant; / be vast; be moth-eaten > waste away; be in dread
> dread; ( adj.) be thirsty; be consecrated; ( adj.) be small; ( adj. )
be slight; be decayed; /
( adj.) be near > draw near; be much;
(adj. )be far; ( adj.) be hungry; ( adj. ) be evil; /
(adj. )
be
sated; /
be bereaved;
(adj.) be desolated; /
(adj.) be glad > rejoice;
(ptc.) hate;
(adj. )
be low.
199
qatal they have a present temporal reference in 54 percent of the cases (326
times out of 606 occurrences) versus a past temporal reference in the other 46
percent (279 occurrences). More importantly, when the discourse context is
delimited by examining only instances of direct speech, which have a separate deictic center from the surrounding discourse (see Miller 2003: 131), the
statistics shift even more favorably toward a default present temporal reference
for stative verbs in qatal: 78 percent (227 of 290 occurrence) show a present
temporal reference versus 22 percent with past temporal reference.
The significance of this patterning in BH is that it is typical of stative verbs
with perfective aspect and past tense conjugations in other languages (so Bybee, Perkins, and Pagliuca 1994: 92; e.g., Schuh 2007: 606, with reference to
Bades completive verb). Thus, not only does it confirm the above dynamicstative interpretation of the a-i/u theme vowel distinction, it presents a strong
typological argument that qatal is a perfective verb form and wayyiqtol is
pasttense.
3.2.3. Viewpoint Aspect: The Perfective-Imperfective Opposition
and the Progressive
Most of the time when the term aspect comes up in discussions of the
BHVS, viewpoint aspect is what is intended. As discussed at some length in
1.3.3 and 1.7.3.3, viewpoint aspect has to do with the way in which language
can encode different viewpoints on a situation, distinct from the temporal constituency (situation aspect) and temporal location (tense) of the situation. The
debate over viewpoint aspect in the BHVS has traditionally been a two-way
debate between tense and aspect, and more recently a three-way argument
among tense, aspect, and mood or modality. Traditionally, this debate has centered on how to characterize the opposition between qatal and yiqtol. The reason for this is that, despite other disagreements over the BHVS, most scholars
still concur with the traditional view that qatal and yiqtol form a central (if not
the central) opposition within the BHVS. This position is fairly self-evident
when one considers the restrictions placed on the other verb forms: wayyiqtol is largely restricted to prose narrative; the Jussive-Imperative constitutes a
directive-volitive subsystem of irrealis mood in the BHVS; and the Participle,
as an adjectival predicate, is not fully integrated into the finite verbal system.
By contrast, neither qatal nor yiqtol is restricted to certain types of literature
(e.g., prose versus poetry), or types of discourse (e.g., narrative versus direct speech), or even certain genres (e.g., story, wisdom sayings, prophetic
speech,etc.).
The few scholars who demure from the traditional view that the qatal: yiqtol
opposition is central to the BHVS sometimes posit a more abstract opposition
that includes a number of verb forms. For example, Joosten (1999: 16) characterizes the system as constituting a primary opposition between indicative
200
Chapter 3
and nonindicative modality (for summary and critique, see 2.4.2.3). Others
object that the entire tense-versus-aspect debate is problematic because these
parameters cannot be distinguished adequately in any TAM system (e.g., Buth
1992). Although there is some validity to these remarks, which I addressed
above (3.1.1), by and large this sort of criticism is simply unhelpful because
it represents a retreat from meaning. This results in an unclassified taxonomy
of meanings/functions for the BH verb forms that is useless to the philological
task for which it was presumably developed (see the especially egregious case
of Andrason 2010; 2011a; 2011b; 2011c; forthcoming). This objection must
in part be rejected when it entails entirely banning questions about the precise
semantic identity of individual grams, such as qatal and yiqtol.
I am defending here the traditional views that qatal and yiqtol form a central
opposition of the TAM system of BH and that this opposition is best described
as perfective (qatal) and imperfective (yiqtol) viewpoint aspect. Granting this
position (substantiated below), it is legitimate and helpful to employ Bhats
(1999) terminology of the aspect-prominence of the BHVS. Not only this
but, while the focus of this book remains the semantics of the individual grams
that make up the BHVS (of which there are tense grams, aspect grams, and
mood grams), identifying the BHVS as aspect-prominent and the central opposition of qatal : yiqtol as perfective : imperfective viewpoint is useful from the
perspective of making typological comparisons and arguments. For instance,
the presence of a distinct class of stative verbs in BH (see above) and the
absence of tense shifting (see Endo 1996: 300) aligns the BHVS with aspectprominent TAM systems versus tense-prominent systems (Bhat 1999: 40, 150).
From the perspective of diachronic typology, the BHVS appears more
aspect-prominent than mood-prominent. Not only are TAM systems with a
primary realis : irrealis (cf. Joostens indicative : nonindicative/modal opposition) opposition rare in the worlds languages (Bybee, Perkins, and Pagliuca
1994: 23738), they usually develop into non-future : future tenseprominent
systems (so Bhat 1999: 17; see also Comrie 1985: 50). This is in contrast to
Hebrew, about which there is general agreement that in the postbiblical period
it developed into a past : non-past opposition, tense-prominent system (Segal
1927: 150).
Complementary to these negative typological arguments, several arguments
in favor of an aspect-prominent analysis have already been touched on and can
be rehearsed here as a general segue to the detailed analyses of the qatal and
yiqtol conjugations and the predicatively used Participle. First, a TAM system
with a core binary opposition of perfective : imperfective viewpoint aspect is
eminently probable based on the extensive typological studies of Dahl (1985)
and Bybee (1985; also Bybee, Perkins, and Pagliuca 1994) and their combined
study (Bybee and Dahl 1989). Bybee and Dahl (1989: 83) note that TAM
systems with a core binary opposition of perfective : imperfective viewpoint
aspect constitutes the most frequent type of TAM system in their data and
201
seems to occur in about every second language in the world (see fig.2.6b in
2.4.2.3, p.138).
Second, buttressing these findings are the arguments of Bybee (1985) and
Dahl (1989) regarding the basic-ness of aspectual oppositions versus tensed
oppositions, which I have already rehearsed in 2.4.2.3 in my critique of
Kuryowiczs theory of the Semitic verb. The integration of TAM morphology with the discontinuous-root morphology of the Semitic (and BH) verb in
contrast to the periphrastic means of denoting tense in some Semitic languages
(e.g., Classical Arabic Past Imperfective with kana yaktubu he was writing,
cited in example [2.15], p.139)34 make it much more plausible that the qatal:
yiqtol opposition is one of viewpoint aspect and not tense or mood.
Third, as already discussed in the previous section (3.2.2), the pattern of
interaction of qatal versus wayyiqtol with stative roots is mutually confirming
of the dynamic-stative analysis of the a-i/u theme-vowel distinction as well as
a perfective analysis of qatal and past tense analysis of wayyiqtol (see below).
Bybee, Perkins, and Pagliuca (1994: 9192) note that this interaction with
stative roots is an important means of distinguishing perfective aspect verbs
and past tense verbs, the semantics of which are closely intertwined (e.g., Dahl
[1985: 79] characterizes past tense as a secondary meaning of perfective
verbs).
Fourth, despite the intermingling of meanings between past tense and
perfective aspect, the predominate past temporal reference for qatal can be
accounted for within an aspectual analysis in light of Smiths analysis of default temporal interpretation of aspectual forms (C.S. Smith 2006; 2008; see
1.7.6 and 3.4.2). Fifth, and finally, as will be demonstrated in the following
subsections, the viewpoint-aspect analysis of the qatal: yiqtol opposition is
consonant with the diachronic data and contributes to a coherent and typologically feasible TAM system for BH.
3.2.3.1.Perfective Qatal
A helpful starting point to my argument that qatal is a perfective aspect
conjugation (i.e., a perfective gram) is the summary of the typical taxonomy
of meanings for the form given in the standard grammars: the qatal may express (1) present or past state (with statives), (2) simple past, (3) past perfect,
(4) present perfect, (5) present (gnomic), (6) performative, (7) future perfect,
(8)counterfactual, (9) so-called prophetic perfect, and (10) optative/precative
(see Bergstrsser 1962: 2.2529; Davidson 1901: 5863; S.R. Driver 1998:
1326; Gibson 1994: 6070; Joon 2006: 112; Kautzsch 1910: 30913;
34.By contrast, and more similar to Hebrew, Aramaic developed periphrastic progressive past using the participle rather than imperfective finite verb form (for Syriac, see
Nldeke 2001: 21618; and Muraoka 1997b: 68; for Babylonian Jewish Aramaic, see Khan
2007: 114).
202
Chapter 3
Waltke and OConnor 1990: 48695).35 The range of temporal reference evident in this taxonomy forms the strongest objection to a past tense analysis of
qatal,36 because it is difficult to explain the full range of temporal reference for
a form, the general meaning of which is identified as past tense. Indeed, Bybee,
Perkins, and Pagliuca (1994: 95) note that a distinguishing feature between
perfective verbs and past tense is the ability of the former to appear with present or future temporal reference. Attempts to explain the perfect meaning for
qatal based on its syntactic context are unsuccessful (e.g., Zevit 1998; see my
critique above, 2.4.2.2), and recourse to Reichenbachs relative tense theory
is problematic inasmuch as Reichenbach ignored the aspectual dimensions of
the perfect (see Comrie 1976: 52; and above, 1.7.3.3).
Theories that identify the qatal conjugation as realis or indicative mood
versus irrealis or nonindicative conjugations likewise falter in dealing with the
basic taxonomy of meanings for the conjugation. In particular, such a theory
cannot incorporate counterfactual examples such as example [3.9]. 37
[3.9] Gen 26:10
One of our people might have lain with your wife so that you
would have brought guilt upon us.
203
the face of clear examples, such as [3.10], of qatals advancing the reference
time (see 1.5; for additional criticism, see Cook 2004b; and chap.4).38
[3.10] Gen 4:18
Irad was born to Enoch, and Irad begat Mehujael, and Mehujael begat Methusael, and Methusael begat Lamech.
In contrast to these unsuccessful approaches, in the remainder of this section I make my case for perfective qatal based first on a diachronic-typological
examination of the form, followed by a discussion of the way that the various
indicative meanings of the conjugation relate to its general perfective meaning
(see 3.1.1 for justification of this approach). Part of this argument will entail
identifying morphologically the so-called wqatal conjugation with qatal, but
since (as I argue) this is a syntactically signaled realis : irrealis distinction
within the qatal conjugation (see 3.3.1 below), I am choosing to treat the irrealis meanings of the conjugation separately below (see 3.3.3).
The development of the BH qatal conjugation is relatively undisputed in
Semitic studies (see 2.3.2.12). Its origin is identified as a Common Semitic verbal adjective *qatil.39 As evidenced in Akkadian, this verbal adjective
came to be combined with the encliticized subject pronoun to express nullcopula predications (e.g., *qarib anta > *qarib-ta you are drawn near; see
von Soden 1995: 77; Kuryowicz 1972: 6465; Huehnergard 2005: 22.1;
Kouwenberg 2010: 7).40 This construction became further grammaticalized
in West Semitic, where the subject pronouns were reduced to inflectional suffixes and the theme vowel of the pattern was changed from the stative i-class
to a dynamic a-class (*qarib-ta > *qarabta; Bergstrsser 1983: 11 notes;
Diakonoff 1988: 90; Huehnergard 1992: 156; Lipiski 1997: 341; Moscati
1980: 133; Tropper 1998a: 182).41 At the other end of the spectrum is the fact
that the verbal system of Hebrew underwent dramatic changes in configuration
38. Examples of this sort, though not frequent, can quite easily be multiplied: Gen 10:24;
17:20; 29:34; 30:8; Deut 5:27; 1Chr 1:18; 2:4849; 2Chr 7:12.
39. Note, however, that a dynamic perfect qatal is posited for East Semitic Eblaite: e.g.,
a-k-al-ma-lik Malik has devoured; da-na-il Il has judged (Mller 1984: 15758).
40. This predicative use of the Verbal Adjective in Akkadian (as Huehnergard 2005 refers to it) is referred to as the Stative (e.g., Borger 1979: 170; see Huehnergard 1987; 1988;
Kouwenberg 2000 for discussion; the older term Permansive has largely been abandoned).
41. Like the stative: dynamic distinction in BH (see 3.2.2), the semantic distinction
between the *qati/ula and *qatala patterns was not consistently observed in West Semitic:
Mller (1983: 38) claims that originally the stative *qatila pattern was not limited to present
time reference or to an active or passive sense (similarly, G.R. Driver 1936: 80); both verb
patterns occur with a dynamic meaning in Eblaite (Mller 1984: 15758); and some roots in
EA occur in both patterns (Rainey 1996b: 303).
204
Chapter 3
West Semitic
(e.g., El Amarna)
Biblical Hebrew
Rabbinic Hebrew
(e.g., Mishnah)
*qarabta
qrat
qrat
between the end of the biblical writings (ca. 2nd century b.c.e.) and the end of
the Tannaitic period of rabbinic literature (ca. 200 c.e., the date generally assigned to the Mishnah). Most obvious is the loss of the wayyiqtol form, which
affected the distribution of the qatal conjugation, and thus the form became
generally restricted to past temporal reference.42 Thus, we can reconstruct the
qatal conjugation as in table 3.1.
While this reconstruction of the form of the qatal conjugation is widely
agreed on, a case must be made for the TAM meaning of the conjugation at
each stage. One approach to this task would be to analyze each relevant occurrence by way of translation into a natural metalanguage (e.g., English), compile statistics, and identify the conjugations meaning as that which appears
in the majority of instances. I reject this sort of approach (see above, 3.1.2)
because it lacks an objective means of validation. A more promising tack is
to ask what type of development of verb forms can be seen in the typological
data that would best explain the development of this form. The starting point
for this comparison is the major characteristics of the conjugation rehearsed
above: (1)the origin of the form is a verbal adjective used as the predicate of
a null copula; (2)the most basic functions of the form in the Hebrew Bible
are to denote perfective, perfect, and past events (see taxonomies of standard
grammars); (3)by the end of the Tannaitic period, the conjugation becomes
restricted to past temporal reference, even with stative roots. Even more importantly, the patterning of qatal with stative predicates in BH and RH strongly
disposes one to identify the TAM of qatal as perfective aspect in BH and past
tense in RH. Given all of these data, the most relevant path of development
is the perfective/past development proposed by Bybee, Perkins, and Pagliuca
(1994: 105) and provided in fig. 3.6.
Variations within this path of development may occur at either end. At the
front end, verb forms may originate in either a resultative (a state brought
42. There are isolated instances in which wayyiqtol in the SamuelKings source text is
replaced with qatal in the Late BH narratives of Chronicles: e.g., 1Kgs 8:4 and 2Chr 5:5;
2Sam 24:4 and 1Chr 21:4; 1Kgs 5:1 and 1Chr 1:26; 1Kgs 12:16 and 2Chr 10:16 (corrupt
text); 1Kgs 15:13 and 2Chr 15:16; 2Kgs 8:27 and 2Chr 22:3; 2Kgs 15:5 and 2Chr 26:20;
2Kgs 16:17 and 2Chr 28:16). These may be evidence of qatals shift from perfective to
simple past (see Senz-Badillos 1993: 120; Polzin 1976: 57; see Kienast 2001: 315). However, there are also passages in which the opposite phenomenon is observable (e.g., 1Kgs
14:21 and 2Chr 12:13; 1Kgs 22:41 and 2Chr 20:31; 2Kgs 18:4 and 2Chr 31:1).
205
resultative (be/have)
perfect > perfective/simple past
completive (finish)
Figure 3.6. Grammaticalization paths for perfective/simple past verbs
(adapted from Bybee, Perkins, and Pagliuca 1994: 105).
206
Chapter 3
developed from a resultative (versus a completive) construction (i.e., *qaribat You are drawn near). T. D. Andersen (2000: 34) has recently objected
to a resultative origin for qatal, arguing that in Proto-Semitic . . . *qatala
anta would have meant you (are) killed, with the subject as patient, not
agent. It is unclear how this could have evolved into *qatalta meaning you
have killed (Andersen 2000: 34). The confusion in Andersens objection
aside (How could *qatala ever have had a passive sense?)44resultatives are
more restricted than perfects, and thus *qatala would have combined initially
only with intransitive predicates (which Andersen apparently confuses with
passive voice).45 Presumably then, the development of the dynamic *qatala
conjugation in Semitic originally occurred with intransitive verbs and subsequently spread to transitives (see p.197 n. 29). This explanation accords with
Huehnergards discussion (2005: 4.3; see also Kouwenberg 2010: 16574)
of the predicative use of the Verbal Adjective in Akkadian, which has a passive sense with transitive verbs and a resultative meaning with intransitive
verbs. It is also consonant with Haugs (2008) recent discussion of this shift
in ancient Greek. Haug (2008: 29495) argues that the crucial step in the
resultative> perfect shift is the inferential identification of the subject of the
main verb with the agent of the participle, which becomes conventionalized
so that the construction spreads to other predicate types besides those to which
the resultative is confined. Haug (2008: 295302) makes his case for the correctness of this surmise for ancient Greek from the development of the Greek
Perfect with middle and active voices. Despite the lack of similar evidence,
we can reasonably hypothesize that a similar sort of conventionalization and
spread from intransitive predicates to all predicates accompanied the themevowel shift from *qatil-ta to *qatalta and was coincident with the reanalysis
of the construction as a finite, dynamic verb form (versus enclitic pronoun, null
copula, and adjectival complement).
Second, and proceeding from the first point, we would expect West Semitic
*qatala to exhibit a primary perfect meaning, and so Daniels (in Bergstrsser
1983: 21 note ac) surmises. Data in support of this contention may be fewer
44. T. D. Andersen (2000: 34) also confuses resultative and perfect in his discussion,
claiming that the functional equivalent in English of the Semitic resultative construction is
have + Past Participle. However, it is clear from Bybee, Perkins, and Pagliucas discussion
(1994: 6367) that resultatives in English are constructed of be + Past Participle (e.g., He
is gone). After identifying the Dravidian Present Perfect as a perfect, Andersen (2000: 41)
translates the form with an English resultative construction: The lights are turned on. This
confusion perhaps accounts for his objection to a resultative > perfect shift for transitive
verbs in Semitic alongside his proposal that *qatala tended to have a resultative meaning
with achievement and accomplishment verbs, which are mostly transitive (T.D. Andersen
2000: 4142).
45. Haug (2008: 294) states it this way: Resultatives are lexically restricted to verbs
with defined resultant state, whereas Anteriors [= perfects] can in general be formed from
any verbs.
207
I say, you and Ziba will divide the field.
b. Gen 28:15
Indeed, I will not abandon you until (the time) when I will
have done that which I have told you.
46. Examples of qatal with a future perfect sense are plentiful in the grammars: e.g.,
Gen 43:14; Num 32:17; Deut 8:10; 1Sam 20:22; Isa 11:9; Jer 8:3; 29:14, 18; Amos 9:15;
Mic 5:2; Ruth 2:21; 1Chr 14:15, etc. Performative qatal is likewise well recognized in the
grammars: e.g., Gen 14:22; 22:16; Deut 8:19; Ruth 4:9; 2Sam 17:11; 19:30; 24:2223; Song
2:7; etc.
208
Chapter 3
West Semitic
(e.g., EA)
resultative
> perfect
*qariba anta > *qarib-ta *qarabta
Biblical Hebrew
Rabbinic Hebrew
(e.g., Mishnah)
> perfective
qrat
The performative example in [3.11a] in particular provides an additional typological argument: while languages are not limited to a specific verb form
for performative expressions, the choice of verb to express performatives is
not haphazard; rather, in languages where tense is grammaticalized, a presenttense form is used (e.g., English), but in languages where aspect is grammaticalized, perfective aspect grams are used, as in BH (see Dobbs-Allsopp
20047: 52, 5758).
Fourth, based again on the distinctions between perfective and simple past
verbs given in Bybee, Perkins, and Pagliuca (1994: 9195), we can identify
qatal in post-BH (i.e., the Mishnah and beyond) as a simple past verb. The
data that have already been discussed above (3.2.2, especially n.31) are the
mirror image of the data for BH qatal; namely, RH qatal is restricted to past
temporal reference, and it does not exhibit a distinctive present-stative meaning with stative predicates. The disappearance of wayyiqtol at the same time as
this perfective > simple past shift of qatal both explains the shift and buttresses
the semantic identification of the form in the post-BH period. 47
Thus, to conclude this section, I propose the reconstruction of forms and
TAM of Hebrew qatal given in table 3.2. This reconstruction has been argued,
not based on statistical analysis of individual examples, but primarily on the
basis of typological arguments along with the consensus views regarding the
primary meanings of the conjugation in each period.
A final issue in this reconstruction of qatal that remains to be addressed
is the status of the so-called wqatal conjugation. Since at least the medieval grammarians, wqatal has been treated as a separate conjugation from
qatal, based on semantic, morphological, and morphophonemic differences
between the two forms: semantically, the meaning of wqatal parallels more
closely yiqtol than qatal, analogous to the semantic parallel apparent between
qatal and wayyiqtol; morphologically, the waw conjugation appears obliga47. The expression of perfect by qatal in RH (see Prez Fernndez 1992: 11617; e.g.,
if the pillar of dawn has not risen in m. Ber. 1:1) cautions against too
simplistic a picture of the verbal system in RH. However, this feature does not overturn the
weight of the other evidence for past tense qatal in RH, and further exploration of the RH
verbal system is beyond the scope of this book.
209
Aside from other misunderstandings in Andersens argument,48 the most obvious objection is that there is no evidence that (West Semitic) *qatala at
any of its stages was limited to particular types of predicate (e.g., transitive,
intransitive).
The morphophonemic argument is perhaps the most well-established argument for a distinct wqatal conjugation. However, it faces several problems.
48. Andersens analysis contains numerous misunderstandings and inaccuracies, from
his confusion of resultative and perfect (see n. 44) and claim that *qatala is a Proto-Semitic
form (a marginal view at best; see 2.3) to his classifying stative under imperfective aspect (2000: 28). He also misreads Bybee, Perkins, and Pagliuca when he claims based on
their data that perfective and simple past verbs rarely function in conditional statementsin
contrast to the frequent use of wqatal in conditional clauses (Andersen 2000: 38; citing Bybee, Perkins, and Pagliuca 1994: 207). Andersen has overlooked the conditional functions of
old anteriors and perfectives listed elsewhere in Bybee, Perkins, and Pagliuca (1994: 79, 93).
210
Chapter 3
First, the ultimate stress on wqatal forms is limited within the paradigm (only
1s and 2ms exhibit the variation, the other forms already showing ultimate
stress in qatal). Second, even for these forms, there is a good degree of variation so that many apparently wqatal forms (on the basis of the waw conjunction and nonperfective meaning) that should show ultimate stress do not, even
within the immediate context of others that do show the expected stress pattern
(e.g., Amos 1:4,
and Amos 1:5,
).
As a result, scholars have
questioned whether this stress pattern could even be formulated as a morphophonemic rule (e.g., C.H. Gordon 1938; Blake 1944a; Sheehan 1970).
Revells examination (1984; 1985) of the stress pattern on the wqatal represents an important argument against the morphophonemic interpretation.
Revell examines the variation in the presence or absence of the ultimate stress
and concludes that it is subject to prosodic forces: Stress position in the perfect forms with waw consecutive is, then, conditioned by the intonation patterns characteristic of speech units into which the text was divided according to
the syntactic, (other) semantic, and rhythmic factors described (Revell 1985:
299; see p.280). More importantly, he argues that the distinguishing mark of
the semantic category waw consecutive perfect almost certainly arose after
this form had ceased to be used even in contemporary literature, and probably
arose within the biblical reading tradition (Revell 1984: 440). Thus, it represents at most a retrospective grammatical analysis and not an organic development within BH grammar.
In conclusion, there is no evidence that qatal and wqatal represent distinct
and independent developments in Hebrew, and therefore I am eschewing the
continued use of wqatal along with the misunderstandings it conveys. Nevertheless, the perceived semantic and syntactic distinctions between qatal and
wqatal are undeniable, and it may be possible even to argue that they are
separate conjugations if one can demonstrate how wqatal branched off from
qatal (the reverse direction of divergent development is unlikely enough not to
merit entertaining it) and that the two are now (in BH) sufficiently distinct. Below (3.3), I argue that the difference between qatal and the so-called wqatal
is a syntactically signaled mood distinction: qatal as realis mood, wqatal as
irrealis mood.
What follows is not intended simply as a taxonomy of meaningsmeanings that are already available in the standard grammars. Rather, it is further
support for the argument that BH qatal is marked for perfective aspect. To support my argument, I pose the question at every step in the taxonomy whether
the particular meaning for the qatal conjugation can persuasively be explained
with reference to a basic meaning of perfective aspect. As with the earlier part
of my argument, here too, we have recourse to typological data: to the extent
that the range of meanings for qatal lines up with the range of meanings gen-
211
In Hebron he reigned seven years and six months and in Jerusalem he reigned thirty-three years over all Israel and Judah.
The question remains how to explain the close association between these two
notions, given that perfective qatal is largely confined to past temporal reference. To argue that, in some instances, perfective aspect is more salient and, in
other cases, past tense is better is simply to sidestep the issue or deny that there
is any such notion as basic meaning in terms of TAM forms. Dahl (1985:
79) concluded from the correlation of perfective aspect and past temporal reference that past tense is simple a secondary meaning of perfective verbs.
While perhaps accurate, his claim fails to explain why this is so. More promising is C.S. Smith (2006: 92; 2008: 35), who argues that in languages with
aspectual conjugations there is a default temporal interpretation assigned to
the verbs based on the parameter of boundedness (on boundedness, see 1.5.3
and chap.4): unbounded situations are located in the present, and bounded
situations are located in the past. Because perfective aspect is one way to effect boundedness, the temporal location of perfective events is constrained by
this default pattern, resulting in the close association between the notions of
perfective aspect and simple past tense. This default temporal interpretation is
central to explaining the expression of temporality in BH (see 3.4.2).
The stative interpretation of stative predicates with perfective qatal has already been discussed at some length in this and the previous section, where it is
shown that this interaction differs from the interpretation of stative predicates
with simple past verbs. Because stative verbs are unbounded, the default in
temporal interpretation is present (see example [3.13a]). However, within pasttime accounts, the unbounded state is delimited to the past by the temporal
deixis of the narrative discourse, thereby yielding a past stative meaning (see
example [3.13b]; see additional discussion in chap.4).
212
Chapter 3
He said to them, I am old, I have advanced in days.
b. Gen 48:10
They eyes of Israel were heavy from old age; he could not
see.
The perfect (or anterior) meanings for perfective qatal were accounted
for above (this section) as the persistence of the conjugations earlier perfect meaning. As with the temporal distinction of stative predicates with qatal
(present or past), the temporal interpretation of perfect verbs is determined by
the deictic characteristics of the discourse in which they appear.
[3.14] a. Deut 1:10
Yhwh your God has increased you, and look, you
(are) today as the stars of the heavens with regard to
numerousness.
b. Gen 5:24
Enoch walked with God and he was no more, because God
had taken him.
c. 1Chr 14:15
When you hear the sound of marching at the tops of
the baca trees, then you shall go out into battle, for God
will have gone out before you to strike the camp of the
Philistines.
For instance, in example [3.14a], the present perfect interpretation is determined in part by the deictic connection to the speaker (as opposed to anchoring
to a previous narrative event) of direct speech and in part by the explication
of the resultant state of the qatal event in a verbless clause that defaults for
the present deictic center of the speaker. Similarly, the past perfect interpretation of qatal in [3.14b] can be explained by reference to the narrative deictic
center of the immediate context (i.e., the preceding wayyiqtol form), which
determines a past temporal interpretation of the resultant state (he was not)
of the qatal, resulting in a past perfect interpretation of the event. Finally, the
213
future perfect interpretation in [3.14c] is, again, determined by the deictic character of the context. Here the temporal adverbial with the directive yiqtol
expressing Gods command to David places the deictic center in the future so
that the perfective qatal, expressing a bounded event, is interpreted as future
perfect. These are just brief illustrations of the way that the temporal character
of a given discourse context determines the precise interpretation of perfective
qatal. I provide a much more detailed discussion in chap.4, after the semantic
groundwork of the default temporal interpretation has been explicated in 3.4,
below.
Of the other meanings listed in the summary of standard taxonomies at
the beginning of this section, the counterfactual and optative/precative are
discussed below under the category of irrealis mood (3.3.3.1). Three other
categories of realis mood qatal remain to be discussed here: the performative use, the employment of the form in gnomic expressions, and the category
traditionally referred to as prophetic perfect because of its association with
prophetic oracular literature.
The performative use of qatal has been introduced above in example
[3.11a] as evidence that it is a perfective gram and is further illustrated by example [3.15a] below: the performative use of this gram demonstrates that the
conjunction may be used with a non-past temporal reference, and the regular
choice of this verb instead of the yiqtol, for example, for performative points
to the crucial characteristic of the gram in this function to be its instantaneousnessthat is, its perfective aspect value. Other present- or future-time expressions with qatal represent extensions of the performative, from which they
are usually not distinguished (see S.R. Driver 1998: 17; Joon 2006: 112g;
Waltke and OConnor 1990: 489). The commissive, by which we commit
ourselves to do things (Searle 1983: 166), is distinct from the performative
only by a perceived temporal gap between the statement and action, as illustrated by example [3.15b]. Waltke and OConnors (1990: 489) commissive
analysis (they label it perfect of resolve) of the performative example [3.15a]
illustrates the close similarity between these meanings. Finally, the epistolary
use of the qatal, illustrated in [3.15c], differs from performative and commissive by yet another temporal gapthe gap between the writing and the reading
of the correspondence.49 The semantics of the qatal form remains constant in
all three examples, namely, [CRF, S & PFV & DECL] (i.e., present temporal
reference, perfective aspect, and declarative mood); the discourse-pragmatic
context disambiguates them.
49. Other commissive examples include Gen 23:13; 17:20; Judg 1:2; 2Chr 12:5. Another epistolary example is 2Chr 2:12 (for others, see the epigraphic examples in Pardee
1983).
214
Chapter 3
. . .
Ephron answered Abraham, The field I (hereby) give to
you, and the cave that is in it to you I (hereby) give it in the
sight of my people.
b. Gen 17:20
As for Ishmael, I have heard you: behold, I (promise to)
bless him, with the result that I will make him fruitful and
will multiply him very greatly.
c. 1Kgs 15:19
Herewith, I send you a gift of silver and gold.
An accurate assessment of qatal in gnomic (better termed generic) expressions requires clarifying because it is still a poorly understood category
(for more details, see Cook 2005, on which the following discussion is based).
First, generics do not refer to particular situations but are probability statements characterizing the properties of certain states of affairs. For this reason,
they are often referred to as omnitemporal. Second, languages do not use any
particular linguistic means of marking generics; therefore, their interpretation
is often ambiguous, both with respect to a generic or nongeneric expression
and with respect to different sorts of generic descriptions (e.g., habitual, descriptive, normative). Third, languages tend to use a minimally marked verb
form or some form of present/imperfective verb in generic expressions. For
this reason, traditional treatments of generic expressions in the Hebrew Bible
have advocated translating them with general present tense verbs (e.g., S.R.
Driver 1998: 17). Thus, the use of BH qatal in generic expressions presents a
challenge both for scholars who argue that the form is a simple past (e.g., Rogland 2003) and for my argument here that the form is perfective aspect. However, as both Rogland (2003) and I (Cook 2005) have show in some measure,
the appearance of qatal in generic expressions is not nearly as prevalent as is
sometimes supposed (see Cook 2005: 124), and it is too often misunderstood.
I have expressly argued (Cook 2005), based on the data of Proverbs, that TAM
distinctions are retained within generic expressions in BH (see Dahl [1985:
100], who makes a general claim to this effect).
Thus, the qatal form in examples such as [3.16] should not be analyzed as
having a present or omnitemporal reference but should be given the sort of
215
A wise man went up to a city of mighty men and brought
down (its) strong fortification.
No harm will befall the righteous, but the wicked are full
of evil.
b. Prov 20:12
An ear that listens and an eye that seesYhwh has made
both of them.
In cases of alternation between qatal and yiqtol in Proverbs, the usual TAM
distinctions of the conjugations are retained, as illustrated by the examples in
[3.18].
[3.18] a. Prov 14:1
Wisdom has built her house, but folly will tear it down
with her own hands.
b. Prov 6:8
It will prepare/prepares its food in the summer, it has gathered its produce in harvest.
The same holds for some examples in Psalms, as in example [3.19a]. However,
other examples appear to fit Helds explanation (1962); he argued that the yiqtol forms in cases such as [3.19b] represent the prefixed simple past conjugation (>wayyiqtol) and not imperfective yiqtol (see further 3.4.1).
216
Chapter 3
[3.19] a. Ps 2:2
The category prophetic perfect, which has been used to describe examples
[3.20ad], among others (see C.L. Klein 1990; Rogland 2003: 53113), clearly
lies in the margins of the verbal system (to borrow Hendels phrase [1996]).
[3.20] a. Isa 5:13
Assuredly, My people will suffer exile for not giving heed,
its multitude victims of hunger and its masses parched with
thirst. (njps)
b. Isa 11:9
In all of My sacred mount nothing evil or vile shall be
done; for the land shall be filled with devotion to the Lord
as water covers the sea. (njps)
c. Isa 19:7
And the Nile papyrus by the Nile-side and everything sown
by the Nile shall wither, blow away, and vanish. (njps)
d. Jer 48:41
Kerioth shall be captured and the strongholds shall be
seized. In that day, the heart of Moabs warriors shall be
like the heart of a woman in travail. (njps)
C.L. Kleins (1990) description of the category makes it suspiciously ad hoc.
He insists that a prophetic statement must be textually clear, in realis (indicative) mood, and have a clear future temporal reference in order to qualify as
a prophetic perfect. Likewise, my (Cook 2002: 22223) explanation in terms
217
Behold, we perish, we are undone, we are all undone; from Kautzsch 1910: 106n) is
more a matter of rhetoric than grammatical tense. That is, stating Behold, we have perished,
we have been undone, we have all been undone while one is still alive (to be able to say it)
would assure that there was no confusion for the hearer regarding the time of death, but it
precludes all possibility of escape from it.
218
Chapter 3
The raiding party departed from the camp of the Philistines
in three companies: one company was turning the way of
Oprah towards the land of Shual; another company was
turning the way of Beth-horon; and another company was
turning the border road that overlooks the valley of Zeboim
in the direction of the wilderness.
b. 1Sam 22:23 (present imperfective yiqtol)
Stay with me; do not be afraid, for whoever is seeking my
life is seeking your life.
51. Other past imperfective examples include: Gen 6:4; Exod 8:20; 19:19; Judg 9:38;
1Sam 1:10; 2Sam 15:37; 23:10; 1Kgs 6:8; 20:33; Isa 1:21; Hos 2:1. As Joostens discussion (1999) illustrates, distinguishing past imperfective (realis mood) from habitual (irrealis mood) can be difficult (e.g., Isa 1:21 and Hos 2:1 both might be past habitual or past
imperfective). Other present imperfective examples include: Gen 16:8; 37:15; Josh 9:8; etc.
In Bybee, Perkins, and Pagliucas view (1994: 126), all present-time indicative yiqtol examples would be by definition imperfective, but this seems to overstate the case (see 1.7.6).
219
Joosten (1999; 2002) accepts the premise that these meanings for yiqtol
are at variance with his nonvolitive modal identification; hence, he attempts
to explain them as nonindicative modal meanings. The individual faults in
his argument aside (see Cook 2006a), his overall approach has two weaknesses. First, in order to explain examples such as [3.21] as modal, he must
make the category modal endlessly elastic, such as by the claim that there
is something inherently modal about questions (Joosten 2002: 54). If yiqtol
is a modal gram, the precise extent of modal should be more clearly delineated than he has hitherto done. Second, given his admission that yiqtol was at
an earlier stage an imperfective (Joosten 2002: 63) or present-future (Joosten
1989: 156; 2007: 58), his theory needs to account more clearly for the way that
this form became modalbeyond simply claiming that the change is common
(so Joosten 2007: 58).
The weakness of the non-past tense identification is even more serious insofar as the examples of yiqtol with past temporal reference are inherently at odds
with identifying the form as an opposite, tensed gram. Zevits claim (1988: 30)
that examples of preterite yiqtol (< *yaqtul) in the biblical text would short
circuit the verbal system is equally valid with respect to the non-past gram
identification of yiqtol and yiqtols with past temporal reference.
A somewhat different objection to the imperfective identification of yiqtol
is the argument of Waltke and OConnor (1990: 47677) that, as the unmarked
member in privative opposition with perfective qatal, it is better labeled nonperfective. Essentially, Waltke and OConnor have reclothed earlier aorist
identifications of yiqtol in the garb of markedness theory:
In this form the notions of aspect and time both blend (imperfective aspect in
past and present time) and separate (aorist in future time). Sperber [1943; 1966]
and Hughes [1955; 1970] are partially right in describing it as a universal tense.
And it may signify more than a blending of tense and aspect or pure tense; it may
also signify either real or unreal moodsthe indicative as well as degrees of
dubiety and volition. In short: a form that can signify any time, any mood, and
imperfective aspect (but not perfective), is not imperfective but nonperfective, a
more than opposite of the suffix [qatal] conjugation. (The term aorist, meaning without limits or boundaries, is not inappropriate.) (1990: 47677)
Insofar as yiqtol is clearly less marked than qatal in its aspectual opposition,
Waltke and OConnor are justified. However, given my goal to determine an
appropriate cross-linguistically comparable gram identity for yiqtol, nonperfective is essentially equivalent to imperfective, understanding (again) that we
are not thereby limiting the form to expressing only imperfective aspect (e.g.,
excluding tense as part of its amalgam of meaning).
The second complication for substantiating imperfective yiqtol is that the
comparative-historical evidence is simply not as transparent for the form compared with the case for qatal. Nevertheless, even given such constraints, diachronic typology is able to shed light on the development and identity of yiqtol.
220
Chapter 3
221
gressive construction that develops into an imperfective gram (Dahl 1985: 93;
Bybee, Perkins, and Pagliuca 1994: 141).53 These two grams are distinguished
primarily by the more generalized meaning of the imperfective, which includes
progressive expressions along with other meanings such as habitual and gnomic (Comrie 1976: 25; Bybee, Perkins, and Pagliuca 1994: 141; Heine and
Kuteva 2002: 93).
However, this progressive-imperfective path of development does not
adequately account for the future and irrealis mood meanings that the yiqtol conjugation expresses in BH (Andrason 2010: 36 n. 50). In this regard,
Haspelmath 1998 represents an important extension of Bybee, Perkins, and
Pagliucas study (1994), which Andrason (2010) applies to BH as a corrective,
by way of extension, to my earlier ideas (Cook 2002). Haspelmath surveys
a number of anomalous grams, such as indicative forms that also express
subjunctive mood, futures that may denote present, and grams with both future and habitual meanings and concludes that they can only be explained as
originating in old presents. That is, the miscellaneous set of functions
(Bybee, Perkins, and Pagliuca 1994: 277) associated with these forms derives
from older grams along the progressive-imperfective diachrony that through
competition with newer progressive grams have lost their central imperfective/
present function (on overlap of imperfective and present, see Bybee, Perkins,
and Pagliuca 1994: 126). These miscellaneous functions, such as future and
habitual, regularly appear as peripheral meanings of imperfective grams (see
Bybee, Perkins, and Pagliuca 1994: 158). But Haspelmaths crucial insight is
that, through competition with other grams, these peripheral and seemingly
miscellaneous meanings may come to be the more dominant meanings alongside the diminished or lost prototypical imperfective meanings. A shift of this
sort would neatly explain the distribution of meanings associated with the BH
yiqtol: because of competition from the participial progressive gram, imperfective yiqtol only infrequently expresses past and present imperfective as residual meanings of its earlier prototypical functions (Haspelmath 1998: 36),
while (general) future and subjunctive meanings are becoming primary functions of the form. This reconstruction removes the objection that yiqtol cannot
be an imperfective gram because it expresses general (perfective) future, in
that such grams are anomalous in part because of the tendency of future and
subjunctive expressions to be perfective (Haspelmath 1998: 55; for a similar
case in Tigre, see Bybee, Perkins, and Pagliuca 1994:14647). This sort of
proposal fits well with the Rabbinic Hebrew data, which point to irrealiss
becoming even more predominantly the central meaning for yiqtol (see Prez
Fernndez 1997: 108).
53. More strongly, Haspelmath (1998: 34 n. 6) states that he knows of no present grams
(a subpart of imperfective according to Bybee, Perkins, and Paglicua [1994: 141]) that
derive from any source other than progressives.
222
Chapter 3
Biblical Hebrew
> imperfective yiqtol
Rabbinic Hebrew
> irrealis yiqtol
223
Wealth will not avail on the day of wrath, but righteousness will deliver from death.
b. Prov 13:20 (Qere)
Whoever walks with the wise will become wise, and whoever consorts with fools will suffer harm.
c. Qoh 3:14
I know that all that God does, it will be forever.
I will multiply the seed of David my servant and the Levites,
my servants/who serve me/the ones serving me.
17:2, 20; 19:5, 9; 21:11; 27:18; 28:18; 29:23; from Qoheleth: 1:3 (or general present), 18
[2]; 2:3 (or general present), 16, 21 (or future); 3:14 [2], 15, 17; 4:10 [2], 11, 12; 5:9
(stative), 11, 17 (or general present); 6:7 (stative), 12; 7:3, 7 [2], 9, 12, 18, 19, 20 [2], 26
[2]; 8:1 [2], 3 [2], 5 [2], 12, 13, 15; 9:4, 11; 10:1 [2], 8 [2], 10:9 [2], 12, 14, 15, 18
[2], 19 [2], 20 [2]; 11:3 [2], 4 [2], 5.
55. The discussion in this section is based on Cook 2008a.
224
Chapter 3
Although this is the only such example in the Hebrew Bible, a fact that makes
the text suspect to some scholars,56 there are plenty of examples of plural construct Participles followed by constituents that are semantically complements
(e.g., Gen 3:15,
knowing good and evil; Josh 3:14, 15, 17,
those lifting the ark). The verbal-nominal ambiguity in example [3.23]
is analogous to the verbal-nominal split in the stative form, as illustrated by
examples [3.24ac].
[3.24] a. Josh 23:2b
He said to them, I am/have become old, I have advanced
in days.
b. Gen 18:11
Abraham and Sarah (were) old, advanced in days; the
manner of women had ceased to be for Sarah.
c. Gen 24:1
Now Abraham was old/(was) old, he had/(was) advanced
in years.
I submit that the ambiguities described above place both the Participle and the
stative alike in the class of adjective, which has long held an uncertain position
between verbal and nominal classes (see Cook 2008a: 56).
At the same time, the Participle when used predicatively overlaps semantically with yiqtol, even in close juxtapositioning, as in example [3.25] (see
Waltke and OConnor 1990: 624; cf. S.R. Driver 1998: 27).
[3.25] Gen 37:1516
The man asked him, What are you looking for? He said, I
(am) looking for my brothers.
Thus, while it is incorrect to classify the Participle as a finite verb form (e.g.,
Joosten 1989: 128), it is likewise unwarranted to exclude the Participle from
discussions of the BH TAM, as previous studies have done (e.g., Ewald 1879).
Examining the status of the adjectival Participle (and stative) in light of
typological data on predicate encoding strategies makes it possible to clarify
56. Joon (2006: 121k n. 2) suggests that either the text is in error and we should read
my servants, as found in the previous verse (vs. 21), or if the text is correct then the
construction is analogous with instances of the construct participle before prepositions (see
Joon 2006: 129m), especially given that the two particles (object marker and preposition
)are often confused in the book of Jeremiah.
225
the precise character of these forms in relation to the TAM system of BH and
also to develop strong implicational typological arguments with regard to the
diachronic development of this TAM system. Stassens (1997: 12) typology of
intransitive predication encodings, listed in examples [3.26ad], provides an
appropriate framework for this discussion.
[3.26] Predicate categories of intransitive predication (Stassen
1997:12):
a. John walks (event/verbal)
b. John is a carpenter (class/nominal)
c. John is in the kitchen (location)
d. John is tall (property/adjectival)
Stassen finds that the first three of these categoriesevent, class, and location
[3.26ac]have cross-linguistically valid prototypical encoding strategies,
described as follows.
First, event predicates are prototypically encoded according to the verbal
strategy [3.26a]. Stassen includes both states and dynamic event types in this
category. The verbal strategy is non-supported; that is, there is no copula
or auxiliary required. In addition, if the language has person-agreement features, the verbal strategy will have person-agreement marking (Stassen 1997:
50, 120). Second, the nominal strategy, which prototypically encodes classmembership predicates [3.26b], in contrast to the verbal strategy, is supported
by a nonverbal or zero copula. That is, either the predicate is introduced by an
overt copula that does not have the characteristics of a verb, such as a pronoun
or particle, or the predicate is juxtaposed with the subject, and the copula is
merely implied (Stassen 1997: 12021). Although English does not exhibit
these sorts of copula strategies, examples [3.27ac] illustrate the strategies.
[3.27] a. Zero-copula (Stassen 1997: 62)
Moskva gorod (Russian)
Moscow city
Moscow (is) a city
b. Pronominal copula (Glinert 1989: 189)
gift pro clock
The clock is a gift
c. Particle copula (Stassen 1997: 79)
Rm pw mw (Egyptian)
man cop Mahu
Mahu is a man
226
Chapter 3
227
Prototypical adjectives or properties (i.e., other than the Participle and stative)
use the same copula strategies as nouns and locative expressions, as illustrated
in example [3.29].58
[3.29] a. Gen 29:31 (zero copula)
But Rachel (was) barren.
b. Lam 1:18 (pronominal copula)
Yhwh is righteous.
c. Gen 11:30 (verbal copula)
Now Sarai was barren.
By contrast, the stative and the Participle exhibit unique and distinct patterns of
predicate encoding. First, the stative allows either verbal or nominal-locative
predicate encodings, as illustrated by examples [3.30ac].
[3.30] a. Gen 48:10 (verbal encoding)
They eyes of Israel were heavy from old age.
b. Exod 17:12 (nominal encoding with zero copula)
Moses hands (were) heavy.
c. Gen 50:9 (locative encoding with verbal copula)
The camp was very large.
Second, the predicative Participle is consistently encoded with nominallocative strategies; however, it is distinct from prototypical adjectives (which
are also consistently encoded with nominal-locative strategies) in that the Participle can encode an event predicate, as in examples [3.31ad].
[3.31] a. 1Kgs 13:1 (zero copula)
Jeroboam (was) standing beside the altar.
b. Deut 31:3 (pronominal copula)
Yhwh your God is passing over before you.
58. Neither [3.29b] nor [3.31b] is as compelling an example of the pronominal copula as
example [3.28b], with its lack of personperson agreement. This sort of uncertainty does not
affect the larger argument I am making here, however.
228
Chapter 3
229
Biblical Hebrew
Rabbinic Hebrew
split nominal-locative
encoding >
hu haya holek ~ hu holek
he was/(is) walking
230
Chapter 3
Finally, these findings form the basis for an important implicational typological argument based on Stassens Tensedness Parameter (1997: 34757),
which states that languages that are aspect-prominent tend to encode adjectival predicates according to their verbal strategy, whereas tense-prominent languages tend to encode adjectival predicates according to one or more of their
nominal strategies. Reversing this implication, the shift away the split verbalnominal stative encoding and the development nominal encodingboth of
prototypical adjectives and of event predicates in the Participleindicates that
the drift of Hebrew is from more aspect-prominent to more tense-prominent.
Identification of the predicative Participle construction as a progressive
gram is compelling on the basis of several pieces of evidence. First, given
the above arguments that the yiqtol is a gram developing on the progressiveimperfective path and given the degree of overlap between it and the Participle
(see example [3.25] above), it makes good sense to see the predicative Participle as a younger (progressive) gram developing along the same progressiveimperfective path. Second, Stassen (1997: 242) notes that all such relationships
of the sort I have posited for yiqtol and the predicative Participle (which he
terms verb switching) take the form of a contrast between simple verb
forms and periphrastic (or complex) verb forms. This datum lends support to
the view espoused above, that the Participle is always copularly supported. It
also confirms the progressive analysis of the predicative Participle in that periphrastic expressions are a major lexical source of progressive grams (Bybee,
Perkins, and Pagliuca 1994: 12733).
As a progressive gram, the predicative Participle expresses an agent in the
midst of an activity at the reference time (so Bybee, Perkins, and Pagliuca
1994: 136), which is determined by the discourse context in which the Participle occurs, whether by lexical modification, as in [3.32ab], by a neighboring
TAM form, as in [3.32cd], or by the narrative reference time more generally
[3.32d].
auxiliary has. These include: (1) change of semantics of the main verb; (2) variation regarding whether or not the construction is used with all verbs in the language; (3) integration of
the structure into the system of paradigmatic distinctions of the language; (4) possibility/impossibility of using adverbials between the auxiliary and main verb; (5) particular facts about
gender/number agreement between subject and object and the components of the auxiliary
structure, etc. Points (2) and (3) and possibly (4) argue in favor of an auxiliary analysis of
in compound tenses in Modern Hebrew. Dyk and Talstra (1999: 165) attempt to nuance
the case in Biblical Hebrew by arguing that a difference should be maintained between stating that the formal syntax allows for a participle to be interpreted as being the main verbal
predication and stating that in a particular instance it actually was intended as the main verbal
predication. However, it is unclear what they mean by actually was intended, nor does
their next comment seem to follow logically: Obviously, something would be missed if one
were to treat the participle merely as a nominal form in examples such as Jeremiah 7:17.
231
Thus says Yhwh: By this you will know that I am Yhwh:
Look, I (am) striking with the staff that is in my hand the
waters that are in the Nile so that they turn to blood.
e. Gen 24:30
And look, he (was) standing by the camels beside the
spring.
In example [3.33], all of the above-listed elements contribute to a correct pastperfect progressive interpretation of the the participle, [CRF < Cpos1 < S &
PROGACT & DECL]: the preceding qatal forms set the past-temporal context;65
the temporal phrase
sets a temporal boundary in the past before
which the situation was taking place; finally, the overt qatal copula reinforces the past-in-the-past (i.e., past perfect) location of the event. 66
65. Note that a series of qatal forms is employed intentionally to avoid the narrative
implication of temporal succession associated with wayyiqtol: Hezekiah did each of these
things as part of his reform but not necessarily in this order (see further chap.4).
66. Examples such as this illustrate well the manner of the development of the present
parameter in terms of verbal copular support of the participle (see above in this section): the
overt copula here clarifies the past-perfect progressive interpretation that otherwise would be
ambiguous between a past-perfect and a simple past interpretation. Such ambiguity appears
in two other places in the text:
is ambiguous between past or past perfect in
Hebrew just as in the English rendering; similarly, and supporting the ambiguous understanding of the preceding, the wayyiqtol
continuation from the past-perfect progressive
participle maintains the continued sense of past-perfect temporal interpretation. Probably a
consequential sense is the main motivation for its use here: because they had been burning
incense to the bronze serpent, they had taken to calling it by a proper name, thus personifying
232
Chapter 3
A slanderer reveals confidences, but a trustworthy spirit
conceals a word.
233
234
Chapter 3
235
constructions, such as apodictic/conditional law codes. This idea fits well with
Bybees (1998: 265) characterization of grams that typically are labeled irrealis as denoting situations that are not asserted and frequently derive their meaning from the construction in which they occur. Thus, and third, the category is
appropriate because it designates the origin of these grams in modal constructions and their dependence on those constructions to inform their meaning.
Finally, the negative argument is Occams razor: if the various nonindicative meanings cohere in a single form and can be reasonably related to that
form in terms of a single general meaning and context-determined individual meanings, this is preferable to positing several different grams that
happen to have the same grammaticalized form. The case is analogous with
the case for allomorphs: if the various phonological shapes are all related to
a general meaning, and the variation can be explained based on phonological context, then it is expedient to posit a morph that represents, albeit as an
abstraction, the various allomorphs as a single morphological entity.
3.3.1. Word Order in BH: An Overview
Since the word-order indicated mood distinction encompasses both mood
systems (i.e., irrealis and directive-volitive), I begin with an overview of word
order in BH. These arguments are not my own, however, but are adopted from
colleagues whose work on word order I am endorsing here. Word order necessarily comes into play in any analysis of the BHVS because the waw-prefixed
conjugations (according to the traditional understanding of them) are almost
always clause initial. Variation among these and the qatal and yiqtol forms
have been explained by some scholars as being due to this syntactic restriction
and not any semantic difference (see 2.4.2.2). Several other theories surveyed
in chap.2 included attention to word order for understanding the semantics of
the BHVS (see 2.4.2.3).
Assessments of word order in BH illustrate a sort of tyranny of the majority in two respects: the overwhelming majority of scholars have held that BH
syntax is basically a verb-subject (VS) order because this is statistically the
most frequent order. However, given that the waw-prefixed forms represent
a syntactically constrained clause type and that wayyiqtol in particular is additionally restricted to past narrative genre, there should be some healthy suspicion about whether frequency alone is sufficient for determining basic word
order in BH. The tyranny of the majority is seen also in the fact that the VS
analysis is often assumed, without any defense, to be the very foundation of
studies of the pragmatics of word order (e.g., Shimasaki 2002, and the criticisms in Cook 2003a and Holmstedt 2003; Moshavi 2010 and the reaction in
Holmstedt 2011).
Suspicions regarding this majority position have been growing and with
good reason: SV languages outnumber VS languages by almost six to one
236
Chapter 3
237
PQD
KBD
QWM
[word order] as a whole must be delayed until a full study of all the texts has
been completed. Yet at this point the case for an SV-realis: VS-irrealis mood
opposition makes the best sense of the word order data generally and, as argued below, the BHVS specifically. For this reason, I proceed on the basis of
Holmstedts SV/X-VS theory, drawing particularly on the word-order mood
opposition included in his theory. This pattern is justified by the statistical correlation between the morphologically distinct directive-volitive irrealis mood
forms and VS order (above) and by claims that the VS-restricted wqatal is
essentially an irrealis mood form (Joosten 1989). Here I extend these ideas and
argue that, system wide, the BHVS displays a mood-based word order distinction, and that the adoption of this position particularly illuminates the position
of wqatal with respect to the rest of the BHVS.
3.3.2. Directive-Volitive Mood System
There are good morphological, syntactic, and semantic grounds for treating the Imperative, Jussive, and Cohortative forms as constituting a distinct
directive-volitive subsystem of BH irrealis mood. Morphologically, the three
forms are constructed on the prefix pattern, which they share in common
with yiqtol and wayyiqtol.71 Thus these conjugations all share a common set of
agreement features, and the vowel pattern in the root-and-pattern morphology
is consistent throughout the conjugations for each verb (see table 3.5).
However, the Imperative, which is limited to the second person, lacks the
agreement prefixes, retaining only gender-number-indicating suffixes (cf.
Imperative mp
Guard! versus yiqtol or Jussive 2mp
You shall
guard). Cross-linguistically, this is similar to some non-pro-drop languages
that do not require an overt subject for the imperative (e.g., English Go!; see
Verstraete 2007: 39).
The Jussive is distinct from yiqtol only with certain roots, verbs, and inflectional forms: in the 3ms and 2ms forms (i.e., forms without any inflectional
71. The final vowel that distinguishes the yiqtol (< *yaqtulu) has been discussed above
(2.3; 3.2.3.2), and the case for a distinct wayyiqtol conjugation is discussed further below
(3.4.1).
238
Chapter 3
sufformative) in (1) the Hiphil binyan, where the Jussive has a different theme
vowel (e.g., Jussive 3ms Let him congregate versus yiqtol 3ms He
will congregate), or in (2) roots with a glide (w or y) in the second or third
position, in which cases the Jussive has an apocopated singular (e.g., Jussive
3ms Let him build versus yiqtol 3ms
He will build; see Kummerow
2007a).
The Cohortative has traditionally been described as a first-person volitive,
distinguished from first-person yiqtol by its paragogic -, which, however,
does not occur on roots with a glide (/ )or glottal ( )in third position or
with enclitic object pronouns.72 Older grammars treat the paragogic - as a
lengthening of the Imperfect, in contrast to Jussive, which is a shortening of
the same (e.g., Kautzsch 1910: 129). The paragogic - has been understood
as contributing some sort of intensification (Ewald 1879: 17) or emphasis
(S.R. Driver 1998: 51) to the verbal action. Moran (1950; 1960, repr. 2003),
followed by Rainey (1975; 1986), suggested that BH Cohortative is a reflex
of a fuller Canaanite Volitive conjugation *yaqtula (see Joon 2006: 117c).
However, the modal interpretation of the paragogic - is questionable on several grounds.
First, the EA evidence for a Canaanite Volitive conjugation is equivocal.
Rainey (1996b: 262) states that it is abundantly clear that the EA texts have
not given us any conclusive evidence for the existence of a Canaanite yaqtula
pattern. In spite of Morans brilliant mustering of the evidence, it is still possible to argue that the -a suffix is merely the Akkadian ventive. This view
is more strongly stated by Lipiski (2001: 26061): This subjunctive in -a
alternates in the Amarna correspondence with the East Semitic ventive, and its
particular use with expressions of wish or expectation, especially in 1 s, leave
little doubt that the West Semitic subjunctive in -a is but a ventive or allative,
with some distinctive syntactical features.
Second, although the paragogic - appears on the so-called Cohortative
over 500 times,73 it occurs over 300 times on the Imperative, with which as a
modal marker it would seem to be redundant. Shulman has examined the 116
examples of the long imperative form (i.e., those with a paragogic -) in
Genesis through 2Kings and has concluded that the long imperative form
is used where the speaker requests an action directed to himself, an action
done for him/to him/towards him/with him etc. (Shulman 1996: 66). Shulman discovered that in 112 cases a prepositional idea with reference to the
speaker was present (61 times) or implied in the context (51 times), thus re
72. Dallaire (2002: 127 n. 228) notes several exceptions wherein roots with a thirdposition exhibit the paragogic -: -Gen 29:21; Judg 15:1; 2Kgs 19:23; Ps 43:4; Gen 19:8;
Judg 19:24; 2Chr 1:10.
73. The Westminster morphology contains 521 examples of the form out of 2813 firstperson yiqtol forms (of which the Cohortative is treated as a subcategory).
239
240
Chapter 3
Neh 5:7; 8:13; 6:11; 13:9, 11, 17, 19 [2], 21, 22)77 and often have no evident
reflexive sense. A development toward conventionalization of the paragogic
- on first-person forms is confirmed by the Hebrew of the Dead Sea Scrolls,
in which the first-person forms almost always appear with the paragogic (notable exceptions are found in scrolls of the biblical and apocryphal books;
Qimron 1997: 177). This conventionalization of the reflexive paragogic - on
first-person forms may explain its obligatory appearance on first-person directive forms: the attraction of a reflexive sense to a first-person speaker-oriented
(or subjective deontic) gram is logical, and the sense of the morpheme is comparable to the dative of interest (e.g., Classical Greek; see Smyth 1956: 341).78
Finally, some scholars have recently suggested that the paragogic - forms
an energic system with the paragogic and energic forms (Rattray 1992:
4749; Testen 1994; Gentry 1998: 2130).79 Significantly, the latter forms
are associated with the realis yiqtol form. The comparative data of the EA and
Arabic verbal systems suggest the presence of one or more -terminating conjugations (see table 2.7, p.117; Moran 2003: 94; W.Wright 1962: 1.62). The
suggestion that these form an energic system with the paragogic - forms
arises from the complementary distribution of the morphemes: paragogic following consonant-terminating prefix-pattern forms, paragogic following
vowel-terminating prefix-pattern forms, and energic before pronouns on prefix-pattern forms (e.g.,
I will guard,
you [f] will guard, and
he will guard him/us; see chart in Rattray 1992: 48, and reproduced
in Cook 2002: 245).
Testen (1993; 1994) has argued that all three of these morphemes (i.e.,
paragogic -, paragogic , and energic )can be traced back to a Proto-Semitic
*-am/-nim. However, while morphologically related, a unifying semantic
sense for this energic system is difficult to determine. The Proto-Semitic
77. The paragogic - also occurs frequently (12 times) on the root give: Num 8:19;
Judg 6:9; 1Sam 2:28; 2Sam 12:8 (2); Ezek 16:11; Ps 69:12; Qoh 1:17; Dan 9:3; Neh 2:1,
6, 9. Polzin (1976: 54) incorrectly states that only 1 example of paragogic - (i.e., so-called
Cohortative) is found in Chronicles (1Chr 22:5); there are, in fact, 11 examples (1Chr 13:2
[2], 3; 19:13; 21:2, 13; 22:5; 2Chr 1:10 [2]; 18:6; 20:9).
78. Shulman (1996) tries to discern a separate nuance for the paragogic - on first-person
modal yiqtol than on the Imperative form: the forms with paragogic - express uncertainty
and intention, whereas the forms without the paragogic - express commitment, and determination (Shulman 1996: 238). This conclusion is less convincing (not to mention less
clear) than the conclusion she draws with respect to the paragogic - on Imperatives.
79. More specifically, the forms are associated almost exclusively with realis yiqtol: of
the approximately 450 instances of energic ( Williams 1972: 82), only 5 occur on a wayyiqtol form (Judg 15:2; 2Kgs 9:33; Job 31:15; 33:24; Lam 1:13) and 3 with negative Jussives
(2Sam 13:12; Job 9:34; 13:21; Rattray 1992: 48); similarly, only 9 of the 304 examples of
paragogic occur on wayyiqtol forms (Deut 1:22; 4:11; 5:23; Judg 8:1; 11:18; Isa 41:5; Ezek
44:8; Amos 6:3), and none on Jussives (Hoftijzer 1985: 23).
241
*-am/-nim is realized in the Akkadian ventive, which is essentially a directional element that denotes motion or activity in the direction of, or to a point
near, the speaker (Huehnergard 2005: 15.2). Its connection to the so-called
Cohortative - has been outlined above. However, the semantic significance of
the morpheme(s) is uncertain, despite Rattrays claim (1992: 112) of a similar
ventive meaning for them. We still await a full semantic study of the forms
(cf. Hoftijzer 1985; Zewi 1999).
Despite uncertainties, the above arguments as a whole call into question the
validity of treating the first-person directive as a distinct Cohortative form.
Rather, it is preferable to treat it as a first-person Jussive. Syntactically, the
coherence of this system is apparent from the fact that the directive forms
stand in complementary distribution with each other: the Imperative is limited to second-person positive directive and volitive expressions; the Jussive,
although it appears in all three persons, is particularly complementary to the
Imperative in the second person, which is almost exclusively the negativedirective (using a special negator, )counterpart to the positive-directive Imperative.80 Conversely to the restriction of second-person Jussives to negative
expressions, the first-person Jussives are almost entirely positive.81 As such,
the BH directive-volitive system falls into the minority category of having a
minimal and maximal pattern (see van der Auwera, Dobrushina, and Goussev
2005): the Imperative has a unique strategy (no agreement prefixes and limited
to second person), so it is the minimal part of the system; the Jussive appears
in all three persons, so it is the maximal part of the system.
Semantically, these forms are described in the grammars as expressing
imperative, prohibitive (Jussive with negator ), hortatory, permissive, and
admonitory modalities (Bergstrsser 1962: 2.4553; Davidson 1901: 8695;
S.R. Driver 1998: 5069; Joon 2006: 114; Gibson 1994: 8083, 1057;
Kautzsch 1910: 31926; Waltke and OConnor 1990: 56479). Determining
the most felicitous label for the system per se is difficult. On the one hand,
the traditional label volitive (e.g., Joon 1923, 2006; Waltke and OConnor
1990: 565) understates the central use of the system (especially the Imperative
and Jussive) as addressee-oriented modalities (i.e., speakers will is imposed
on the addressee). On the other hand. the typological label of imperativehortative (van der Auwera, Dobrushina, and Goussev 2005: 294) is too narrowly specific, expressly excluding the prohibitive (i.e., negative directive)
80. A search of unambiguous second-person Jussive forms turns up only 3 out of 91
examples lacking an negator (1Sam 10:8; Ezek 3:3; Ps 71:21). The 2 examples preceded
by in the Leningrad Codex (Exod 10:28; Deut 2:9) are in error, judging from manuscript
evidence and context.
81. Of 518 occurrences of first-person Jussives with paragogic -, only 9 are negated
(with negative): 2Sam 24:14; Jer 17:18; 18:18; Ps 25:2; 31:2, 18; 69:15; 71:1.
242
Chapter 3
and optative (i.e., expression of speaker will without any imposition on the
addressee) meanings associated with the system. When we examine this system with respect to the taxonomies in the first chapter, this system straddles
the speaker-oriented and agent-oriented categories in example [1.29]which
is just another fact that highlights the infelicitous character of that taxonomy.
A more useful approach to classifying the system is to adopt Verstraetes category of subjective-deontic modality (2001, 2007). As such, the system encompasses the directive and volitive modalities in my proposed taxonomy in
table 1.19 (p. 71) as directive-volitive. Directive here is not limited to positive
commands but also includes negative, prohibitive expressions. This table is especially helpful because it encompasses both the positive-directive Imperative
and the negative-directive (negated) Jussive grams. Volitive in my classification does not distinguish between expressions that impose the speakers will on
the addressee (hortative) and expressions that do not (optative), because both
senses are expressed by the BH directive-volitive system and are sometimes
difficult to distinguish in a given example.
The association of certain meanings as dominant with one or the other form
in the system is a by-product of the complementary distribution of the forms
among persons and numbers. For example, the second-person Imperative and
negated Jussive predominantly express directive modality, as illustrated in examples [3.38ab]; only rarely does the Imperative express volitive modality,
such as in [3.38c].
[3.38] a. Jonah 1:2
Get up, go to Nineveh, the great city, and proclaim against
her that their wickeness has risen before me.
b. Gen 22:19
And he said, Do not stretch out your hand against the boy,
and do not do to him anything.
c. Ruth 1:9
May Yhwh grant you and may you find:impv.fp rest, each
in the house of a husband.
The first-person Jussive is largely limited to volitive expressions in the singular, because it is difficult to distinguish a reflexive imposition of the speakers
will (directive or hortatory) from the simple expression of a speakers will
(simple volitional). By contrast, in the plural the first-person Jussive frequently
has a directive (hortatory) idea because the addressees are broader than just the
speaker. These are illustrated by the examples in [3.39].
243
[3.39] a. Ps 9:3
I will rejoice and I will exult in you; I will sing about your
name, Elyon!
b. Gen 37:17
The main said, They set out from here, for I heard them
saying, Let us go to Dothan.
And now, please let your servants settle in the land of
Goshen.
d. Ps 25:2
My God, in you I have trusted, I will not be put to shame;
do not let my enemies exult over me.
e. Ps 104:31
244
Chapter 3
A contingent (purpose/result) modal meaning has been attributed to the directive forms in certain syntactic contextsnamely, when prefixed with the waw
conjunction and following a clause with another directive form (or irrealis
yiqtol with a directive meaning). Joon (1923: 290; 2006: 116) introduced
the label indirect volitive (Fr. volitif indirect) for these examples, in contrast
to direct volitives (Fr. volitif direct; see also Niccacci 1990; Gropp 1991;
Dallaire 2002; see Muraoka 1997a: 229 n.6, for a list of others who use these
labels). Shulman and Muraoka have both questioned this contingent meaning
for directives. Shulman (1996: 221) claimed that the implicated meanings in
these cases are contextual. More strongly and clearly, Muraoka (1997a: 240)
concluded his discussion as follows:
In summing up we would say that the syntagm in question does not have a function of normally indicating purpose. A sequence of volitive verb forms is a series
of so many expressions of the speakers or writers wish and will. The fact that
in some cases the second verb can be more elegantly translated as indicating a
purpose of the first is essentially a question of pragmatics and translation techniques, and not of descriptive grammar and syntax.
Thus, example [3.38c] is translated literally above, based simply on the syntactic coordination of the two clauses; however, the relationship between the
clauses implied by their juxtaposed semantics may justify rendering the verse
May the Lord grant that you might find rest. This treatment underscores the
need to distinguish between the syntactic and semantic relationship between
clauses: the waw signals a coordinate syntactic relationship, but semantically
the clauses may be related in terms of contingent modality, as in the case of
the so-called indirect volitive discussed here and frequently with the irrealis
qatal (3.3.3.1).
The directive-volitive system is summarized in table 3.6. Morphologically,
the forms are constructed on the prefix pattern. Syntactically, they are in complementary distribution among three persons and positive and negative expressions. Semantically, there are tendencies among some persons toward mainly
directive (second person) or volitive (first person) expressions.
3.3.3. Realis-Irrealis Mood Opposition
Justification of the term irrealis has been given above (3.3) and the directive mood system, which aligns with irrealis VS word-order, has been examined (3.3.2). Here the focus is the two primary conjugations, qatal and yiqtol,
and the irrealis meanings that they express in accordance with their VS word
order.
3.3.3.1.Irrealis Yiqtol
Grammars recognize the various irrealis mood meanings for yiqtol (see list
of meanings, 3.2.3.2 above); however, explaining how these meanings came
245
First Person
Second Person
Third Person
Prefix Pattern
Positive
Negative
Jussive
(rare)
Imperative
+ Jussive
Jussive
+ Jussive
mostly volitive
mostly directive
directive and volitive
However, the logic of this conclusion escapes me. For one thing, this statement is apropos of yiqtol in Rabbinic Hebrew, but BH yiqtol lies at an earlier
stage of the same diachronic path, in which it is not yet reduced to modal
uses (and thus also not a proper modal diachrony). For another, the very
83. The fact that yiqtol does at times express dynamic aspect without ( e.g., Deut
1:12, example [3.44a], p.248) does not weaken this argument.
246
Chapter 3
But of the fruit of the tree that is in the middle of the
garden God said, You must not eat from it and you must
not touch it lest you die.
However, the distinctive directive negator remains a crucial disambiguating feature between the prohibitive (negative directive) Jussive, illustrated in
example [3.38b] above, and the prohibitive yiqtol, illustrated in [3.41b]. The
semantic distinction between these two prohibitive expressions has resisted a
clear explanation beyond intuitive statements such as made by Shulman (1996:
128, 187), who states that the Jussive has more urgency about it. In a subsequent article, she writes:
247
The difference between utterances, in which these forms [Jussive and yiqtol] occur, is close to the distinction between deontic and epistemic modality. Jussive
forms are typically used for expressing deontic modality (wishes, commands and
other expressions of volition). The indicative forms, although they may be used
for either deontic or epistemic modality, are typically used for epistemic modality. (Shulman 2000: 180)
Shulmans ideas suggest that the difference lies in the speakers commitment
to the obligation: with the Jussive (and the entire directive system), the obligation is speaker originated, whereas the obligation is only described by the
irrealis yiqtol. The latter is thus similar to epistemic judgment in that epistemic
modality only commits the speaker to the proposition in terms of knowledge,
not obligation.
Nevertheless, Shulmans introduction of epistemic modality is somewhat
confusing. Better suited to describing the difference in question is Verstraetes
(2007: 3235) distinction between subjective and objective deontic expressions: subjective deontics have their source of obligation in the speaker,
whereas objective deontics have a source of obligation outside the speaker
whether expressed or left unexpressed. This speaker-oriented source of obligation explains the urgency of the directive forms noted by Shulman (above),
while the association of irrealis yiqtol with epistemic modal expressions,
as in example [3.42], explains its suitability for objective deontic modality,
hinted at by Shulman (i.e., analogous to epistemic judgment, objective deontic
expressions only describe the obligation without making the speaker a source
of this obligation).84
[3.42] Gen 24:39
And I said to my master, Perhaps the woman will not follow
after me.
84. This explanation rooted in the source of obligation is more convincing to me than
Hatavs (2006: 739), who connects the urgency that Shulman observed to the deictic
character of the directive-system forms, which always take the speech-time as their
reference-time.
248
Chapter 3
How can I bear alone your trouble and your burden and
your bickering?
b. Gen 29:2
He looked and, behold, there was a well in a field and,
behold, three flocks of sheep were lying beside it, because
from that well they would water the flocks.
249
Honor your father and your mother in order that your days
might prolong on the land that Yhwh your God is giving to
you.
Given that subordinating words trigger inversion to VS word order, as does
irrealis modality, the two features of the VS order may have mutually fed one
another; that is, irrealis mood marking with VS was induced by the association
of irrealis mood with subordinate constructions, which are X-VS.
3.3.3.2.Irrealis Qatal (Including So-Called Wqatal)
In the section on perfective qatal (3.2.3.1), I presented the negative argument, that there is no evidence that qatal and wqatal are reflexes of distinct
conjugations in Semitic and that at best one might argue that BH wqatal has
developed into a distinct gram from qatal. In the word order discussion above
(3.3.1), I have provided a positive argument, that the VS word order exhibited by wqatalusually prefixed with a clause-boundary waw conjunction
aligns the form with the irrealis yiqtol and directive mood system, suggesting
that it be reanalyzed as irrealis qatal. This reanalysis is not simply a change
in nomenclature, because there is not complete coincidence between the traditional category of wqatal and what I am treating here as irrealis qatal. Rather,
the latter category includes all the instances in which qatal exhibits VS word
order and irrealis mood.86 This includes, notably, the qatal forms that appear
after ,
,
, or in conditional protases, which are not in the wqatal
category because the latter has been defined as including the waw conjunction
as part of the conjugation, just as in the case of wayyiqtol.
tively, in which qatal predominates (see [3.46], p.250). By contrast, -prefixed qatal occurs
6,459 times versus 1,421 -prefixed yiqtol forms.
86. For example, the qatal passages discussed in examples [3.46ab] and [3.9] are all
treated with the (realis) perfective qatal in the standard grammars (e.g., Kautzsch 1910:
106p; Waltke and OConnor 1990: 30.5.4), but here I am classifying them as irrealis qatal
based on their semantics; word order alone is insufficient, since each has a fronted syntactic
element that may account for the VS order apart from irrealis mood.
250
Chapter 3
The main categories of meaning for irrealis qatal include contingent modality (i.e., conditional constructions, broadly defined, and final clauses), directive
modality, and habituality.87 Counterfactual conditional expressions commonly
feature grams with a past-temporal reference on the basis of the past-irrealis
metaphor: that which is temporally removed from the speakers present is irrealis (Bybee, Perkins, and Pagliuca 1994: 79; compare with English Had I
known, I would have helped). In BH counterfactuals, the conditional words
(positive) or ( negative) may be used, as in [3.46], or other general contexts may indicate a counterfactual statement, as in [3.9] (p.202).
[3.46] a. Judg 13:23
His wife said to him, If Yhwh had desired to kill us he
would not have accepted from our hand a burnt offering and
meal offering.
b. Judg 14:18
He said to them, If you had not plowed with my heifer
you would not have discovered my riddle.
Although Bybee, Perkins, and Pagliuca (1994: 79, 93) include in the data
two cases of perfective verbs that are used to express general hypothetical
conditions, this usage is an unexpected development for perfective qatal, in
spite of its unsurprising appearance in counterfactual expressions. As I argued
above, it cannot be attributed to a different gram from qatal; but comparative evidence shows that cognates of BH qatal frequently acquire a non-past
temporal reference in conditional structures in many Semitic languages, including Arabic, Ethiopic, Aramaic and Syriac, Phoenician, and EA Canaanite
(W. Wright 1962: 2.1417; Dillmann 1974: 548; Folmer 1991; 1995: 394
415; Nldeke 2001: 2035; Krahmalkov 1986; Moran 2003: 3133; Rainey
1996b: 35565). In these data, non-past time examples of qatal cognates ap87. The existence of an irrealis optative or precative meaning for qatal is even more
questionable than the prophetic perfect meaning (3.2.3.1 above). Note, however, that qatal
can have an optative counterfactual meaning when lexically marked by : Isa 48:18; 63:19.
G.R. Driver (1965: 60; S.R. Driver 1998: 25-26) has examined candidates for optative or
precative qatal in the Hebrew Bible and concluded that on the one hand the optative or
precative use of qtl is theoretically as possible in Hebrew as in the cognate Semitic languages. ... On the other hand, all the supposed instances in the Old Testament are doubtful;
none are unavoidable and all can be otherwise explained. After examining all the passages
he treats, in addition to those referenced by S. R. Driver (1998: 25), I concur with G.R.
Drivers assessment (1965): all the examples can be explained in terms of indicative meanings for qatal or textual problems (e.g., Ps 4:2; 7:7; in Lam 1:21, the Septuagint renders the
qatal as an imperative).
251
252
Chapter 3
context C. Which of the two senses is implied usually is, but need not be, dependent on the relevant communication situation. . . .
Stage II: The existence of a sense B now makes it possible for the relevant
form to be used in new contexts that are compatible with B but rule out sense A.
Stage III: B is conventionalized; it may be said to form a secondary focus
characterized by properties containing elements not present in Awith the effect that F now has two polysemes, A and B, which may develop eventually
into homophones.
If he has no daughter, then you should give his inheritance to
his brothers.
At stage 2, in Heine, Claudi, and Hnnemeyers scheme, these irrealis meanings can be conveyed by VS qatal apart from the overt modal contexts that
gave rise to them. Thus, we find final and directive senses of the irrealis qatal
appearing outside conditional apodoses, as illustrated by examples [3.49ab].
[3.49] a. 2Kgs 5:6
Look, I have sent to you Naaman my servant so that you
might deliver him from his leprosy.
b. Exod 25:8
They should make a sanctuary for me that I may dwell in
their midst.
At the same time, examples [3.46ab] illustrate the shortcomings of the above
hypothesis: the modal context of conditional constructions does not consistently obviate the default past-time interpretation of irrealis qatal. One recent
study (Kaufmann 2005) suggests that the tenseless interpretation of verbal
forms in conditional clauses is incorrect and that the TAM of verbal forms
in conditional structures is identical to their interpretation in isolation. This
suggests that the approach to conditional structures may be as wrong-headed
as the general-present interpretation of generic sentences: the TAM of the BH
verbal forms should be interpreted in conditional structures as they would be
253
If I have sinned, then you have watched me and from my
iniquity you will not acquit me. (compare nrsv If I sin,
you watch me . . .)
b. 1Sam 11:3
And if there is no one who will rescue us, then we will
have surrendered to you.
c. Gen 18:26
Of course, more frequently than these past or past-in-the-future interpretations, the irrealis qatal in the consequent clause expresses a modal directive
sense, as in example [3.48] above. As with the case of irrealis yiqtol versus
the directive-volitive forms in directive expressions, the distinctive nuance of
the irrealis qatal in directive expressions has eluded explanation. Two characteristics appear to be crucial to disambiguating among the Imperative/Jussive,
irrealis yiqtol, and irrealis qatal directive expressions. The first has to do with
the source of the obligation, already outlined above (3.3.3.1.): the directivevolitive forms are limited to expressing subjective deontic modality, in which
the source of obligation derives from the speaker, while irrealis yiqtol directives may locate the source of obligation with the speaker or outside them. The
preference of irrealis qatal for procedural directives and law codes seem to
place it alongside irrealis yiqtols expression of obligation, which derives from
some source other than the speaker. However, the frequent use of irrealis qatal
in conjunction with subjective directive-volitive forms, illustrated in example
[3.51], indicates that irrealis qatal like irrealis yiqtol may be used for subjective
or objective deontic modalities.
254
Chapter 3
Yhwh said to Moses, Get up early in the morning and station
yourself before Pharaoh, and (then) say to him, Thus says
Yhwh the God of the Hebrews, Dismiss my people so that
they may serve me.
The second relevant characteristic is the aspectual distinction between irrealis yiqtol and irrealis qatal that makes the latter preferable for procedural
directivesthat is, directives that require being carried out in a particular order, as illustrated by example [3.52].
[3.52] Exod 25:1014
They should make an ark of acacia woodtwo and a half
cubits its length, a cubit and a half its breadth, and a cubit and
a half its height. And then you should overlay it with pure gold
on its inside, and on its outside you should overlay it. Then you
should make upon it a gold molding round about. Then you
should cast for it four gold rings and place (them) upon its four
feettwo rings on one of its sides and two on the other of its
sides. Then you should make poles of acacia wood and overlay
them with gold and insert the poles into the rings on the sides
of the ark, to lift the ark by them.
This preference for using irrealis qatal for procedural instructions may best be
understood as being grounded in its identity as a perfective gram: perfective
qatal presents events as bounded (see further on boundedness in 4.2.1) and
therefore may effect temporal succession among events ( just as realis qatal
and wayyiqtol effect in temporal succession past narrative texts). Because irrealis mood is tenseless (see Verstraete 2007: 4247), the successiveness
of the bounded perfective irrealis qatal is transferred to the sphere of modal
alternative situations: the accessibility of situations is successive (see 1.6
and 4.3.3).
This hypothesis is in keeping with the view, expressed above, that the TAM
of a verbal gram in isolation is maintained in conditional structures: the argu-
255
Irrealis Qatal
(procedural)
ment here would be that the perfective aspect of the qatal gram is maintained
throughout its uses, for both realis and irrealis mood expressions. Buttressing
this argument is the similar preference for irrealis qatal in procedural habitual
expressions, as illustrated by example [3.53]: notice the singular use of the irrealis yiqtol ( ) in v.2 to express habituality of the larger situation of watering the flocks (cited in example [3.44b] above, p. 248), whereas the discourse
shifts to irrealis qatal to report the habitual procedures for the watering of the
flocks (see further discussion in chap.4).
[3.53] Gen 29:23
He looked and, behold, there was a well in a field and, behold,
three flocks of sheep were lying beside it, because from that
well they would water the flocks: all the flocks would gather
there, then they would roll the stone from upon the mouth of
the well, then they would water the flock, then they would
replace the stone upon the mouth of the well.
By way of conclusion, the irrealis mood forms may be mapped according
to table3.7, which illustrates the complementary and overlapping relationships
among the directive system, irrealis yiqtol, and irrealis qatal.
Directive modality is the primary category shared among directive-volitive
forms, yiqtol, and qatal but, as discussed above, the three differ in their particular nuances: directive-volitive forms express subjective deontic obligation
while irrealis yiqtol and qatal may freely express either subjective or objective
deontic obligation; however, irrealis qatal expresses deontic obligation predominantly in procedural contexts. The volitive use of yiqtol is best understood
256
Chapter 3
257
assigned to wayyiqtol in the standard grammars are in line with these statistics:
(1)simple past (usually with the idea of succession); (2)present perfect and
past perfect (the latter under restricted circumstances); (3) logical consecution (past or present time), (4)some exceptional (apparently) future uses in
prophetic contexts (Bergstrsser 1962: 2.3645; Davidson 1901: 7078; S.R.
Driver 1998: 7099; Gibson 1994: 95102; Joon 1993: 38996; Kautzsch
1910: 32630; Meyer 1992: 2.4446; Waltke and OConnor 1990: 54363).
However, the semantic analysis of wayyiqtol has been complicated by the
morphological similarity of the conjugation with imperfective yiqtol, on the
one hand, and the semantic parallels between the conjugation and perfective
qatal, on the other hand (see 2.1). These seemingly contrary relationships lie
directly behind the long-developed view of the form as expressing some sort
of sequentiality in addition to its semantics, whether this parameter served
to distinguish it from yiqtol (so S.R. Driver 1998: 72m and-he-proceededto-say; see 2.2.3) or from qatal (Joon 2006: 118a). But in the century
since Bauers influential theory (1910), in which he posited two distinct prefix
conjugations in order to account for the semantic distinctions between (among
others) BH wayyiqtol and yiqtol, this explanation has come to hold hegemonic
status in BH studies, though not in the same form as Bauer presented it.89
Rather, the usual explanation now is that two prefix conjugations, which were
originally distinct because of the presence of a final vowel on the one (i.e.,
*yaqtul versus *yaqtulu), partially merged morphologically through the loss
of final short vowels (see 2.3 for the full comparative-historical argument). In
BH, reflexes of the short *yaqtul form appear in the Jussive and the wayyiqtol,
while the long *yaqtulu form is realized in yiqtol. This explanation accounts
for the long-observed morphological parallel between the Jussive and wayyiqtol, wherein they both exhibit the same apocopated pattern when possible (see
discussion of Jussive and prefix pattern in Kummerow 2007a). The question
of the relationship between the Jussive and wayyiqtol reflexes of *yaqtul (i.e.,
whether there are two homonymous *yaqtul forms or a single polysemous
form; Huehnergard 1988: 20) seems both intractable and mootintractable
inasmuch as no clear diachrony between the two reflexes can reasonably be
posited given their distinct semantic contrast; and moot, in that, given the wide
semantic discrepancy between the reflexes in BH, there is no risk of confusing the forms. Nevertheless, two other issues regarding the morphology of
wayyiqtol (which in some respects are just as intractable) require comment
here: the first is the distinctive waC- prefix; the second is the question whether
any reflexes of the preterite *yaqtul form occur in BH without the distinctive
waC- prefix.
89. The hegemony of this explanation is evident in its appearance in introductory grammars (e.g., Seow 1995: 22526), which are generally limited to endorsing widely held grammatical ideas.
258
Chapter 3
The origin and shape of the waC- prefix are a long-standing problem (see
list of proposed explanations in McFall 1982: 21719). Analyses may be classified as one of three types (see Testen 1998: 19394): first, the prefix consists
of the waw conjunction plus some other element that has been assimilated into
the following consonant to create the geminated prefix (e.g., Ewald [1879:
19], who proposed an assimilated then); second, the prefix is an alternate
form of the conjunction with a different meaning (e.g., G.R. Driver [1936: 92],
who suggests that the alternate form is analogous to the Akkadian -ma suffix
on verbs); third, the prefix represents a secondary distinction made simply to
disambiguate wayyiqtol from the waw conjunction plus the imperfective yiqtol
or Jussive (e.g., S.R. Driver 1998: 72; Mller 1991).
Given that the waC- prefix is universally present on the wayyiqtol form in
prose narrative,90 apart from a handful of disputed cases (e.g., in Judg
2:1), it is probably justifiable to identify the prefix as part of the grammatical
form of the BH Past Narrative (so Washburn 1994: 4041). However, at the
same time, this third type of approach simply begs the question why the prefix
has the form that it has in BH. Testen (1998: 217) has proposed a theory of
the first type, that waC- consists of the waw conjunction prefixed to a particle
that originated as an *l but became syllabic (*) in environments without an
adjacent vowel.
Testen proposes that this form is the origin of several particles in Semitic,
including the definite article (haC-), precative la-, asseverative l-, which is
realized in Akkadian as l (e.g., l iqtul may he kill; see von Soden 1995:
81; Huehnergard 2005: 29.3), the negative lam in Arabic (*l + neg. m; e.g.,
lam taf al you did not do; see W.Wright 1962: 2.2223, 41), and the waCprefix in BH wayyiqtol. Testen (1998: 203) traces the haC- definite article and
waC- prefix in Hebrew to a common source by analogy with the allomorphs of
the definite article in Arabic: (a)l-/(a)C-. He hypothesizes that in Arabic and
Canaanite the syllabic * developed into aC- (as seen in Arabic), but because
Canaanite did not allow initial vowels, a paragogic was preposed to the
particle. In the case of waC-, that has elided with the addition of the conjunction, just as in the case of the haC- definite article with a preposed preposition
(e.g., , babbyit in the house < *b- + haC- + byit).
Testens etymological explanation, that waC- preserves a particle or function word within the geminated prefix, is preferable syntactically to the other
sorts of explanations because it can account for the obligatory VS order for
wayyiqtol in terms of triggered inversion (so DeCaen 1995: 128; Holmstedt
2009: 125; see 3.3.1).91 Semantically, the analysis of the waC- prefix is even
90. The question whether the conjugation appears in poetry without the waC- prefix is
a separate issue (on which, see below this section) in that most scholars treat such instances
as archaic pastsi.e., not simply the Past Narrative without the distinctive waC- prefix.
91. DeCaen (1995: 128) also identifies a function word in the prefix waC-, explaining it as a phonologically impoverished (he uses the term underspecified) subordinating
259
She said, Alas, I am a widow; my husband died.
In addition to these examples that run counter to the sequential analysis of
waC-, in 4.2.1 I demonstrate that temporal succession is dependent on a variety of factors, including aspect and adverbial modification, making it a priori
unlikely that the prefix itself indicates temporal succession. We must conclude
either that the semantic value of the waC- prefix eludes us or that the function
word has become semantically bleached.93
Hatav (2004) has recently proposed a semantic explanation for the waCprefix that builds conceptually on Testen (1998) and DeCaen (1995), though
she does not appear to be aware of Testens theory and only makes passing
reference to DeCaens syntactic explanation of the prefix (Hatav 2004: 498
n.14). Taking the same tack as DeCaen, she argues that the prefix conjugation
in wayyiqtol is modal, and similar to Testen, she analogizes the function word
in the waC- prefix with the article to argue that it anchors the event in the real
world; the waw conjunction then adds the sense of sequentiality. The motivations behind her theory are two: first, she is committed to a synchronic explanation in contrast to the comparative-historical proposals that identify the prefix
form as a reflex of a preterite or past tense verb (Hatav 2004: 493); second, she
is convinced that BH is a tenseless language (2004: 523; see Hatav 1997).
Both of her motivating arguments illustrate long-standing misunderstandings
in discussions of the BH verbal system, which I have referred to in 2.4.1.
conjunction. However, DeCaen identifies the verb form to which waC- is prefixed as modal
Jussive in order to account for the VS word order.
92. Compare the parallel passage with [3.54a], cited in example [2.5] (p.81):
because he had heard.
93. Semantic bleaching refers to the loss of an items semantic contenteither partial
or wholewhile its grammatical content is retained (see Hopper and Traugott 1993: 87;
Lessau 1994: 75).
260
Chapter 3
First, the question whether yiqtol and wayyiqtol are reflexes of two distinct
conjugations is not strictly a diachronic question but a synchronic question,
since homonymy by its very definition is a synchronic phenomenon. While
comparative-historical data conveniently present evidence for two conjugations, one could just as easily argue from the distribution of the two forms in
the biblical text that yiqtol and wayyiqtol are distinct conjugations. Second,
Hatav treats TAM as discrete categories, so that, if BH is tenseless, it therefore cannot have any tensed grams. Earlier in this chapter (3.2.3), I adopted
Bhats terminology of aspect-prominent (1999) to avoid just such a conclusion as Hatavs, given my argument that qataland yiqtol form an aspectual
opposition. There are, in fact, aspectual, tensed, and mood grams in BH so, to
speak about it as a tense or aspect or mood system is problematic except as I
have qualified it in terms of the prominence of one or the other parameter.
Few scholars any longer dispute that examples of the prefix conjugation
that appears in wayyiqtol occur without the characteristic waC- in the Hebrew
Bible (but see Zevit 1988), as illustrated in example [3.55] by virtue of the
parallel passage in which the form has the waC- prefix.94
[3.55] 2Sam 22:16
And the channels of the sea appeared; and the foundations of
the world were exposed.
Despite the contrast between the parallel texts of 2Samuel 22 and Psalm 18, it
is inaccurate to think of such examples as simply the wayyiqtol without the prefixed waC- in light of the characterization given above. Instead, it is probably
best to follow the lead of previous scholars in treating examples of this sort as
archaic past forms (e.g., Waltke and OConnor 1990: 31.1.1; Rainey 1986;
Greenstein 1988 on the prefixed preterite), in contrast to the more developed
and restricted Past Narrative wayyiqtol.95 Unfortunately, because these occurrences are fully homonymous with the Jussive (and therefore partially with
yiqtol), identifying them as such is mostly ad hoc and often open to dispute.
Alternatively, scholars have long sought to establish the archaic past in
specific syntagms in which the prefix form prima facie expresses past time:
following ( then, 20) and ( before, 26; e.g., Greenstein 1988: 8;
Waltke and OConnor 1990: 498; Meyer 1992: 3.4344; cf. Rainey 1988: 35).
However, the identification of the archaic preterite in these cases is problematic on several grounds: first, the prefix verb exhibits the expected apocopated
94. Other possible examples of the archaic past (preterite) are Exod 15:56; Deut 2:12;
32:8, 10, 11, 13; Judg 2:1; 2Sam 22:14 (cf. Ps 18:14); Ps 18:12 (cf. 2Sam 22:12); 24:2.
95. Compare with the parallel in Ps 18:16, where the form is wayyiqtol: . The precise discourse-pragmatic role of wayyiqtol in prose and poetry is discussed in chap.4.
261
And to Seth also was born a son and he named him Enosh.
Then (people) began to call on the name of Yhwh.
Second, Rabinowitzs conception of complete versus incomplete action is confused: how can an event that is conceived of as having occurred before another
completed action be conceived of as relatively incomplete with reference to
that complete action?
Rabinowitzs explanation is partially endorsed by Revell (1989: 11) who,
however, reinterprets the approach within his tense model of the BHVS: An
262
Chapter 3
96. Compare with the narrative account expressed as using with qatal:
It happened, before he had finished speaking, look, Rebekah was
coming out (Gen 25:15). In this case, the discourse
( see 4.3.4) sets the narrative deictic center in the past (Cpos1) and the qatal in the past context shifts the time back one step
further (CRF) to express a past-in-the-past (past perfect): [CRF < Cpos1 < S]. The participle,
expressing a progressive event, is then indicated as intersecting the past perfect action by the
adverbial .
263
Thus, the -yiqtol syntagm does not preserve the wayyiqtol past tense form
without the waC- prefix and should not be interpreted as past perfective. Instead, yiqtol in this syntagm in a past context has the sense of a relative future in
the pastthus accounting for the three examples with paragogic or energic.
As mentioned, Hendels approach does not explain the use of yiqtol following . In fact, does not function like but temporally locates the
event as approximately simultaneous with the preceding narrative context (so
Kautzsch 1910: 314). The TAM values of the verb forms following are
not limited by the adverb. Thus, we are back to the initial problem created by
the 8 examples of + yiqtol (i.e., excluding the uncertain 1Kgs 8:1 and its
parallel 2Chr 5:2), for some of which we can find parallel or analogous passages with wayyiqtol (cf. Josh 8:30 and 1Kgs 11:7 with Gen 8:20 and 1Sam
14:35; compare Exod 15:1 and Num 21:17 with Judg 5:1). There is no easy
explanation for these forms, though on the basis of the semantic comparisons
with wayyiqtol, they may tentatively be identified as archaic past, which their
contexts seem to dictate.97
The case for wayyiqtols being a narrative past tense form is discussed
in chap. 4. Here, however, is the place to make the case for a past tense
identification of the gram. As with the rest of the prefix pattern conjugations,
the data do not allow a confident etymological reconstruction; however, the semantic similarity with qatal makes it reasonable to hypothesize that wayyyiqtol
developed likewise along the resultative-perfect-perfective-past diachrony. 98
These verbs relative placement on this path is evident from their contrastive
interaction with stative predicates: qatal with stative predicates can express
past or (default) present states, whereas wayyiqtol consistently exhibits a
97. The orthography in 1Chr 5:2 ( ) may indicate that these examples of the archaic
past have been altered in the process of transmission, being written plene and then reinterpreted as yiqtols.
98. Kouwenberg (2010: 130) and Andrason (2011a: 35) both reason backwards that
the form originated from the combination of prefixed pronouns and an adjectival/participle
form based on the overwhelming evidence that resultatives derive from such constructions
(Bybee, Perkins, and Pagliuca 1994: 6768). Both scholars (Kouwenberg 2010: 129 n.11;
Andrason 2011a: 48) also point out my misreading of the data in Bybee, Perkins, and Pagliuca (1994) and my erroneous *ya-qrub reconstruction in my identification of the origin of
wayyiqtol with the infinitive *q(u)tul (Cook 2001: 130, 134; 2002: 239), similar to Bauers
reconstruction (1910) of yiqtol that I follow (see 3.2.3.2 above). The difference between
the reconstructions of these two grams (i.e., *yaqtul and *yaqtulu) illustrates the difficulty
of being certain about their origins: despite the similarity of form, the two are presumably
differentiated by more than simply the final -u vowel; in the case of *yaqtulu, however, we
are on firmer ground both in terms of the typological data (see Bybee, Perkins, and Pagliuca
1994: 12829) and an identifiable infinitive form *q(u)tul that morphologically resembles
*yaqtulu. By contrast, there is no identifiable past participle in Semitic for Kouwenbergs
and Andrasons reconstruction of *yaqtul, and so Kouwenberg (2010: 130) quite appropriately classifies this reconstruction as a matter of speculating.
264
Chapter 3
past-state meaning with stative predicates (see 3.2.2).99 This divergent behavior with stative predicates establishes the semantic distinction between perfective qatal, which defaults for past temporal expressions (see further 3.4.2)
and past tense wayyiqtol (including also the aptly termed archaic preterite
discussed above). The loss of wayyiqtol in later Hebrew at the same time that
qatal was beginning to express past tense points to the competition between
these two forms: wayyiqtol was the older form on the diachronic path, which
had come to be restricted to a narrative role in BH, and later disappeared from
the TAM system; qatal as the younger form functioned more widely in BH
and showed significant semantic overlap with the narrative wayyiqtol while
maintaining a discourse-pragmatic distinction with it (see chap.4).
The prototypical simple past meaning of wayyiqtol is illustrated in example
[3.59].
[3.59] Josh 19:47
265
And when she saw that he had left his garment in her hand and
fled outside, she called to the men of her house.
The examples presented in grammars of wayyiqtol as expressing irrealis
mood or non-past events are all suspect. Most should be translated as simple
past or perfect (e.g., Isa 2:9; 5:15, 25; 24:18; 44:1213; Jer 4:16; 8:16; Mic
2:13; Joel 2:23; Ps 7:13; 20:9; 64:810; 94:2223; Job 14:17) and/or have
textual problems (e.g., Ps 22:30; 49:15; 109:28). The remaining examples that
appear truly problematic are philological rather than linguistic problems (e.g.,
Ezek 33:4, 6; Ps 37:40; Job 5:1516; 36:7). The fact that there are no indisputable future examples of wayyiqtol is another argument that the form is simple
past, since, according to Bybee, Perkins, and Pagliuca (1994: 95), simple past
grams cannot express future events.
3.4.2. The Default Pattern of Temporal Interpretation in BH
In his discussion of imperfective yiqtol, Rainey (1986: 7) protested against
the conventional wisdom that if the BH verbal forms are aspectual the language is tenseless (e.g., Hatav 2004: 523), noting that the verbal system as a
whole does indicate tense and that the communicational context disambiguates the temporal interpretation. He wryly concluded: The ancient Israelite
farmer certainly knew when to milk his cow and his language was adequate to
explain the routine to his son. In this section as well as more extensively in
101. The issue of wayyiqtol and reference time movement entailed by examples such as
these is treated in chap.4 in terms of subsidiary foregrounds (4.2.3).
266
Chapter 3
chap.4, I attempt to delineate precisely what has too frequently been vaguely
referred to as the contextual temporal interpretation of the BH verbal system.
In this section, I apply C.S. Smiths recent research (2006; 2007; 2008;
Smith and Erbaugh 2005) on temporal expression in tenseless languages
(particularly Mandarin Chinese and Navajo) to BH in order to make a case that
the aspectual grams in the BHVS (i.e., qatal, yiqtol, and the Participle) have a
default temporal interpretation, most clearly evident in reported speech, which
is in keeping with comparable aspectual forms in other languages.
Smith (2006; 2008) argues that three basic principles can account for how
aspectual grams are assigned a default temporal interpretation. The most important of these is the deictic pattern of temporal interpretation, listed in [3.62].
[3.62] The Deictic Pattern of Temporal Interpretation
(C.S. Smith 2006: 92; 2008: 235)
a. Unbounded situations are located in the Present
b. Bounded situations are located in the Past
Boundedness has been discussed in 1.5.3, has been discussed briefly in
3.2.3.1, and is discussed with reference to temporality in discourse in 4.2.1.
Sufficient for the present discussion are the following: first, boundedness refers
to whether an event is portrayed as having reached an endpoint or not (Depraetere 1995: 23);102 second, perfective aspect is one means of making an
event bounded, while imperfective and progressive aspects are strategies for
making an event unbounded. This second point in light of [3.62] explains the
oft-noted observations that perfective aspect and past tense are closely related
(e.g., Dahl 1985: 79) and that imperfectivity and present time are at times
indistinguishable (Bybee, Perkins, and Pagliuca 1994: 126). Smiths second
principle, [3.63], is thus simply a corollary of the first:
[3.63] Bounded Event Constraint (C.S. Smith 2006: 92)
Bounded situations may not be located in the Present
The well-recognized exceptions that prove this rule are reportative speech (see
1.7.6) and performatives, which reach their endpoint during the speech interval (see discussion in 3.2.3.1).
Based on these two principles, we can claim with respect to BH that qatal
and wayyiqtol present bounded events, while yiqtol and the Participle present
unbounded events. Although basically correct, this conclusion requires tempering with Smiths third principle [3.64]:
102. Note the distinction that Depraetere (1995: 23) makes between boundedness and
telicity: the latter refers to whether an event has an inherent or intended endpoint (see
1.5.3).
267
Joab came to the king, and he said, What have you done/
did you do?
b. Deut 1:30
Yhwh your God, who is going before you, he will fight for
you.
268
Chapter 3
c. Gen 37:1516
The man asked him, What are you looking for? He said,
I (am) looking for my brothers.
Note that the promiscuousness of yiqtol and the Participle in the expression
of present and future events is such as to prompt perhaps the default temporal
interpretation given in [3.66]:
[3.66] Default Pattern of Temporal Interpretation in BH
a. Perfective qatal is temporally interpreted as past.
b. Imperfective yiqtol and progressive Participle are temporally interpreted as non-past.
Recently, Matheus (2011) has argued that the deictic difference between
speech and non-speech requires that the BH verbal system be treated separately in each deictic context. While his approach is significant in this respect,
it also overstates this situation in a way that is analogous to the way that traditional wisdom overstates the difference in grammar between prose and poetic
texts: there is not a fundamentally different TAM system at work in speech and
non-speech deictic contexts; rather, the temporal organization of non-speech
discourse is less directly related to the canonical speech time than it is (generally) more complicated than the temporal organization of reported speech. For
this reason, rather than attempting to construct a separate default pattern of
temporal interpretation of the BH verbs in non-speech discourse here, I explore
in chap.4 the temporal organization of various modes of discourse with the
aim of clarifying the semantic and discourse-pragmatic interaction about the
temporal interpretation of the BH TAM.
3.5. The TAM System of BH in Diachronic-Typological Perspective
In the preceding sections, I have presented a semantic theory of the TAM
system of BH from the perspective of diachronic typology. I chose this approach as a means of solving the problem of validation of a semantic theory
that deals with ancient texts for which we have no native speakers (i.e., we
have no sure way of testing for felicitous or grammatical expressions). The
theory may be deemed successful to the degree that I am able to map general
meanings to specific meanings by recourse to interaction of various parameters syntagms. This approach is thus far incomplete, and in chap.4 I extend
it to discourse contexts, with specific examples of the way that an analysis of
the TAM system in specific passages might proceed.
At the beginning of this chapter, I argued that more-extreme approaches
to meaning are unhelpful and are unwarranted by the data (e.g., grammatical-
269
Biblical Hebrew
Post-BH
270
Chapter 3
Perfective (qatal)
Past Narrative
(wayyiqtol)
counterfactual
perfect (all times)
performative
simple past
habitual
contingent
Imperative
& Jussive
directive
volitive
Imperfective (yiqtol)
future-in-the-past
general future
progressive/
imperfective
(past & present)
general
prospective
present
progressive
(future)
Participle
involving one or the other of these domains, this does not confine these grams
to these domains. The basic oppositions treated in this chapter that constitute
the skeleton of the BHVS are stative: dynamic situation aspect, perfective:
imperfective viewpoint aspect, and realis: irrealis mood. The first of these is
crucial mainly for its role in certain key typological arguments that distinguish
perfective and past grams.
The other two distinctions are key to understanding the system as such because they cross-cut one otherthat is, the perfective and imperfective grams
function in both the realis and the irrealis domains. The system is visually captured in fig.3.7. The larger, darker-bordered areas of the Perfective qatal and
Imperfective yiqtol underscore the centrality of their aspectual contrast in the
271
system. At the same time, the smaller circles encompassing the Past Narrative
wayyiqtol, Participle, and Imperative and Jussive forms underscore their distinct but less-central position in the BHVS: the Past Narrative is a specialized
verb form in BH that becomes obsolete in post-BH; the Participle is not a finite
verb but, in a supported copular structure, is a productive progressive gram;
the Imperative and Jussive constitute a distinct directive-volitive system. The
meanings associated with each gram are cross-linguistic semantic categories
that are traditionally associated with TAM systems. Thus, certain distinctions
discussed in this chapter are excluded from this diagram. For example, distinctions among various directive expressions discussed above (3.3.23.3.3)
are not included here; commissive is included as part of performative (both
entail the speakers committing to something) inasmuch as it differs, not semantically, but in terms of whether there is a perceived gap in time between
the speech and the action (see 3.2.3.1 above). Although generic (or gnomic)
expressions prefer null copula, yiqtol, and the Participle, they are not restricted
to these predicates, and in particular generic anecdotal narratives (see example
[3.16]) illustrate that generic is not a semantic category of the same sort as the
TAM meanings included here. The strict focus on semantic categories in this
mode also means it is unable properly to distinguish among the categories of
progressive and imperfective because they differ primarily behaviorallythat
is, the imperfective encompasses a broader array of meanings, is more closely
associated with non-past time and irrealis mood, and so on.
These limitations inherent in this strictly semantic model set the stage for
chap.4, in which I turn my attention to the question of the discourse function
of the BH verbal forms. After investigating the semantic-discourse function
interface, I propose there a taxonomy of temporal discourse modes for BH and
illustrate a semantic analysis of the BHVS with them in the context of various
discourse modes.
Chapter 4
272
273
Thus, in the opinion of discourse analysts, the discourse-pragmatic functions that may be correlated with verbal forms are a priority for understanding
a verbal system. This approach creates three methodological problems. First,
discourse studies have been criticized for too-quickly making the leap from
correlation to causation. Downing (1995: 6) cautions that, [w]hen particular
language structures are used in particular discourse contexts, say, . . . in a passage devoted to storyline development, it is sometimes difficult to determine
whether the relationship between the linguistic form and the discourse factor
is causal or merely correlational (see also Tomlin 1995: 545).
The blurring of this line between correlation and causation is evident
in some of Paul Hoppers discussions: in an early study of aspect and the
foreground-background discourse distinction, Hopper (1982: 15) concluded
from the strong correlation between foregrounding and perfectivity that the
primary function of perfectivity is to foreground events in discourse. The examination of foregrounding and movement of reference time below (4.2)
makes it clear, however, that perfectivity is only one of many features that
contribute to foregrounding events. Slightly differently, Hopper and Thompson
(1980: 251) claimed that frequent employment of transitive constructions in the
foreground of discourse contributes to transitivitys grammatical and semantic prominence. However, DeLancey (1987: 54) demurred, claiming that the
explanation of transitivity should be semantic rather than discourse-pragmatic.
Second, without a semantic component, discourse-pragmatic claims about
verbs are often circular; there is no objective means by which to support or
contest such claims. For instance, Hatav (1997: 21) observes an inherent circularity in Longacres dynamic verb rankings: The main difficulty with this
notion [of dynamic verb ranking] is that it is not defined by objective metalinguistic means, which results in a circular claim (wayyiqtol is a dynamic
form because the situation it denotes is dynamic, and the situation is dynamic
because it is denoted by a dynamic form). Bache (1985: 22) leveled a similar
criticism at Weinrichs discourse approach to European languages:
First of all, the fact that the theory is unassailable (to use Weinrichs own
word) makes it rather suspicious. As the saying goes: a theory which cannot be
mortally endangered cannot be alive. As it stands, Weinrichs theory fails to offer
the rigid set of criteria for determining the validity of its own claims which one
would expect of an unassailable theory. It simply relies on our intuitive ability
to tell discursive communication from narrative communication (of course, independently of tense choice since otherwise the unassailable theory is circular).
Third, and perhaps the underlying problem with discourse analyses of verbal
systems, is that they present (explicitly or implicitly) their discourse-pragmatic
explanations as suitable alternatives to semantic explanations. For instance,
many biblical scholars have been content with identifying wayyiqtol as a sequential narrative form without examining a possible semantic motivation for
274
Chapter 4
In other words, discourse approaches to verbs make valid and helpful observations about how verbs function in discourse; however, their explanations
are inherently circular because of their self-imposed limitation to the realm
of discourse-pragmatics. Therefore, I align myself with Fleischmans view
(1990: 23) on the relationship between semantics and discourse-pragmatics of
verbal systems: The pragmatic functions of tense-aspect categories in narrative are not arbitrary; rather, I see them as motivated extensions of the meanings of those categories, extensions that, according to the view of grammar
as emergent (Hopper 1987) may ultimately contribute to a reshaping of the
basic meanings (see also Comrie 1985: 2629).
Although the preceding critique of discourse analysis may appear harsh, it
is not intended to be dismissive of the discipline (see e.g., OConnor 2002).
Discourse analyses furnish important, even necessary observations regarding
275
276
Chapter 4
(e.g., Marchese 1988). Conflation of the two concepts is accounted for by the
fact that sequential chains are frequently employed in temporally successive
discourse, as illustrated by examples [4.1ab] (see Longacre 1985: 26384).2
[4.1] Sequential Chaining in Selepet (from Longacre 1985: 238)
a. Kawa ari-op
kiap ya taka-op
kawa left:3s rem past patrol officer that arrived:3s rem past
Kawa left.
That patrol officer arrived.
b. Kawa ari-mu kiap ya taka-op
Kawa left and that patrol officer arrived.
When the two statements in [4.1] are expressed separately, as in [4.1a], the
verbs left and arrived are both marked with the suffix -op, indicating thirdperson remote past tense. However, because Selepet is a sequential language,
when these clauses are expressed in a chain, as in [4.1b], only arrived is
inflected for tense, aspect, and mood; left is marked instead by the -mu suffix, which only indicates third-person, subject switch. No current theory of the
BHVS that I am aware of supports an analysis of the waw-prefixed forms as
sequential in this sense (there is no evidence that either of the waw-prefixed
verb forms is in any way underspecified); however, the 19th-century wawinductive theories may be interpreted as positing a sequential analysis (see
2.2.1 and McFall 1982: 246).3
Temporal succession is the natural or default interpretation of narrative
texts, whereby, in the absence of any indicators to the contrary, the first-mentioned event is understood as occurring first, the second-mentioned event second, and so on (see Brown and Yule 1983: 125, 144). In other words, the linear
ordering of events in texts is implicitly understood to mirror the order of their
occurrence in the depicted world (Reinhart 1984: 780). Fleischman (1990:
131), therefore, describes temporal succession as diagrammatically iconic:
An iconic diagram is a systematic arrangement of signs, none of which necessarily resembles its referent in respect to any prominent characteristic, as in
the case with an iconic image; rather, it is the relationship of the signs to one
another that mirrors the relationships of their referents (see Hopper and Traugott 1993: 27; 2003: 27). Diagrammatic iconicity of temporal succession is
illustrated by contrasting examples [4.2ab]. In [4.2a], the events are diagram2. Apropos of the above discussion (4.1), Longacre (1985: 265) warns that correlations
such as these do not invariably warrant a claim of causation.
3. A more convincing sequential syntagm in BH is the stringing together of Infinitives
Absolute headed by a fully inflected verb (e.g., Zech 7:5; see Rubinstein 1952); Gai (1982)
argues that the Infinitives Absolute in such examples in BH and cognate Northwest Semitic
languages (esp.Phoenician) should rather be analyzed as underspecified forms that adopt the
specifications of the lead verb (i.e., sequential).
277
278
Chapter 4
279
280
Chapter 4
activities are also allowed (though not required) to extend beyond the perfective reference frame, but their default interpretation is that they are included in
their reference frame, although again, this implicature may be canceled (C.S.
Smith 1999: 488). These conclusions are illustrated by examples [4.7ab].
[4.7] a. Evan knew the answer yesterday (sta),
. . . and he still knows it today.
. . . but today he does not know it.
b. Tage played with his toys (act) . . . and ate his cookie.
. . . and fell asleep.
Without further context, the state in [4.7a] is interpreted as extending beyond
the perfective RF, and the activity in [4.7b] is interpreted as included in the
perfective RF. The implicatures may be either reinforced by the context (first
clause option), or canceled (second clause option).
In summary, perfective aspect makes accomplishments and achievements
bounded, without the need for adverbial modification, because the inherent
endpoint of these situation types is always included in the scope of the perfective viewpoint. Perfective aspect with activities has an implicit temporal
bound whereby the default interpretation is that the activity is bounded, but this
implicature interpretation may be either reinforced or canceled by the context.
Finally, perfective aspects with states are not bounded and are by default interpreted as extending beyond the perfective scope; however, as with activities,
this default interpretation is due to implicature and may be either reinforced or
canceled by the context.
By contrast, any situation aspect with imperfective aspect (in a neutral context) is unbounded, as illustrated in example [4.8].4 In each case, a subsequent
event (introduced by the asymmetrical and then) in the discourse is understood
as interrupting the interval of the imperfective event, regardless of the situation aspect.
[4.8] a. Evan was playing (act), and then he ate lunch.
b. Tage was building a tower (acc), and then he fell down.
c. Jared was winning the race (ach), and then he tripped.
These examples also serve to illustrate the distinction between (a)telicity and
(un)boundedness, since the accomplishment and achievement events remain
[+telic] while being unbounded by their combination with imperfective aspect;
4. No example is given of a state because it does not generally occur with the English
Progressive (used here to illustrate imperfective aspect), though the combination is common
in colloquial English (e.g., He was loving his job more every day); states are unbounded in
any case (i.e., with perfective or imperfective aspect; see [1.45], p.74).
281
their lack of boundedness gives the interpretation that their inherent endpoint
has not been reached.
As already alluded to above, temporal adverbial expressions (e.g., three
times, from 10 a.m. to noon, for three hours, etc.) can be used lexically to effect boundedness in any situation and viewpoint combination. This adverbial
strategy is illustrated by examples [4.9ad] (based on [4.4] and [4.8] above).
[4.9] a. Colin was happy for a few minutes (sta), and then he
started to cry.
b. Evan was playing all morning (act), and then he ate lunch.
c. Tage was building a tower this morning (acc), and this
afternoon he fell down.
d. Jared was winning the race for 30 seconds (ach), but then
he tripped.
Two other factors affecting temporal succession may be mentioned. The first
is subordination, whereby temporal succession is explicitly canceled through
syntactic rather than semantic means (e.g., Evan cried when he went to bed; see
Heinrichs 1986; example [4.2b] above). The second is discourse-pragmatic:
Dowty (1986: 43) observed that the structure of accomplishments is such that
they always allow the inference of temporally included subevents; similarly,
ter Meulen (1995) labels accomplishments filters to describe the contextually determined choice between a temporally successive interpretation and an
interpretation in which the following events are understood as being temporally included in the accomplishment. Thus, in example [4.10], the accomplishment Bill built a house is understood as consisting of the subsequently
described subevents.
[4.10] Bill built a house (acc).
He drew up plans (acc),
bought the lumber and hardware (acc),
and poured the foundation (acc).
However, the interpretation of accomplishments as including subevents is
based on pragmatic implicature; it is not semantic, since the accomplishment
remains properly bounded. By contrast, since activities are only implicitly
bounded, these temporally included subevents would indicate that the activity is unbounded, as in example [4.7b], first option (cf. example [4.14] below).
In such cases, the accomplishment event is perhaps best characterized as the
discourse topic (see Brown and Yule 1983: 7183).
Thus, in example [4.11], the accomplishment he did what was upright is
interpreted as consisting of several subsequently reported events: removing
the high places, shattering the standing stones, cutting down the Asherah, and
282
Chapter 4
any combination of viewpoint and situation aspects with a lexically asserted temporal overlay (e.g., subordination) [4.2b]
smashing the bronze serpent. Knowledge about the nature of an event such as
he did what was upright (i.e., the event is complex) and the subsequent events
as being considered upright in the discourse context lead to the implicational
reading that the latter events compose the former.
[4.11] 2 Kgs 18:34
283
Figure 4.1. Principle of one-sided contour (adapted from Reinhart 1984: 788).
we do find, that most languages only obligatorily mark departures from ordo
naturalis in discourse (i.e., they mark temporal overlay). Second, based on an
understanding of the constellation of semantic factors that contribute to temporal succession, it is illogical to suppose that temporal succession is determined
solely by a particular verbal conjugation.
4.2.2.Foreground-Background
The concept of a foreground-background distinction is recognized by almost all linguists (but cf. Givn 1987), but defining it has been problematic.
Many linguists and virtually all scholars working on BH narrative state the
distinction almost as if it were self-evident. DeLancey (1987: 65) observes that
some factors involved in the foreground-background distinction are psychological rather than purely linguistic. Similarly, Erbaugh (1987) has produced
psycholinguistic evidence of the universality of foreground-background in oral
narratives. Appropriately then, Reinhart (1984) draws on Gestalt theory to
demonstrate the psychological reality of foreground-background in discourse
(on Gestalt psychology, see p.66 n. 47); the key conclusions from her study
are summarized here.
As already mentioned in 1.7.2, space has long been used as a metaphor
for time, as demonstrated by the many temporal concepts that have developed
from spatial concepts (e.g., before, expressing the spatial idea, is metaphorically extended to express the temporal idea of priority; Heine, Claudi, and
Hnnemeyer 1991: 48). Because of this metaphorical relationship between
space and time, examination of key principles in Gestalt theory is particularly enlightening regarding both the reality and the character of the temporal foreground-background distinction. Reinhart (1984: 787) claims that the
foreground-background distinction is a cardinal principle of the organization
of narrative texts and goes on to argue that this distinction reflects principles
of the spatial organization of the visual field: foreground-background in narrative is analogous with figure-ground in Gestalt theory.
To understand the similarity between foreground-background and figureground, consider fig.4.1. The figure is more readily perceived as a square lying on a rectangle than as a rectangle with a square hole in it. Thus, the square
is the figure, and the rectangle the ground. Figure 4.1 illustrates the Gestalt
284
Chapter 4
Foreground-Background
1. Foreground is functionally dependent on the
background.
2. Foregrounded events
are usually temporally
successive.a
3. Punctual [+telic] events
more easily serve as
foregrounded events.
4. Foregrounded events are
usually perfective aspect.
a. On this principle, I depart from Reinhart (1984: 801), who claims that only temporally successive events are candidates for foreground; I explain my position below, in 4.2.3.
principle of the one-sided function of contour: the contour shapes its inside,
not its outside (Koffka 1935: 181). In other words, the lines of the square
more readily define the inside figure as a square than the outside figure as a
rectangle with a hole. The prominence of foregrounded events, in contrast to
backgrounded events, is analogous to the prominence of the figure over the
ground, explained by this Gestalt principle (Reinhart 1984: 803).
Another important analogy between spatial ground and temporal background is that they are both assumed to continue underneath the figure (spatially) or concurrently with the foregrounded event (temporally), even though
they are not explicitly seen or depicted (Dowty 1986: 59). Reinhart (1984:
78788) explains that, if we imagine fig.4.1 as a square book lying on a rectangular table, we assume that the table top continues behind the book.
Reinhart proceeds from her analogy of foreground-background and figureground to apply the four specific principles of the figure-ground relationship
by analogy to foreground-background, as shown in table 4.2.
Functional dependency is illustrated by the shapes in figs.4.2ab, in which
the figure is characterized as a square or a diamond depending on the ori-
a.
285
b.
Figure 4.2a. Diamond figure; 4.2b. Square figure (adapted from Koffka 1935:
185; Reinhart 1984:789).
entation of the rectangular ground (Koffka 1935: 185; Reinhart 1984: 789).
Functional dependency with respect to the foreground-background relationship explains why the intuitive notion of foreground as the gist of a discourse fails: one cannot simply remove the backgrounded events, since they
are functionally interdependent with the foregrounded events, just as the figure
is functionally interdependent with the ground.
The law of good continuation is illustrated by fig. 4.3. The pattern of a
single white horizontal stripe crossing over top of two black stripes is more
easily perceived than four black boxes, because the white stripe is the more
continuous part of the figure (see Reinhart 1984: 803, for other illustrations).
Analogously, foregrounded events, for which temporal succession is the default interpretation, are more prominent in discourse because of their temporal
continuity (Fleischman 1990: 133).
Reinhart draws an analogy between the Gestalt principle of proximity and
the predominance of punctual (telic) events in the foreground of discourse. The
principle is illustrated in fig.4.4, in which we more readily perceive three thin
stripes or pairs of lines, with one extra line to the right, than three wide stripes
or pairs of lines, with one extra line to the left (i.e., one perceives the smaller
286
Chapter 4
|| || || |
][ ][ ][ ]
spaces between the lines more readily than the larger spaces). Analogously, a
series of punctual events are more salient than durative events.
Finally, Reinhart connects the Gestalt principle of closure with the use of
perfective aspect to denote foregrounded events. This principle is illustrated in
a comparison of fig.4.5 with fig. 4.4 above. In fig. 4.5, the effect of the brackets
is to reverse the areas of figure and ground in fig. 4.4 by enclosing the larger
areas of space between the lines. Similarly, where perfective aspect effects
boundedness, the events are viewed as salient. Reinharts analogy between the
principles of the foreground-background distinction in narrative discourse and
figure-ground distinction in Gestalt theory support the psychological reality of
the distinction as well as explain certain regular correlations between aspect
and foreground or background.
4.2.3. The Relationship between
Temporal Succession and Foreground
The preceding discussions (4.2.1 and 4.2.2) may give the impression that
foreground is coterminous with temporal succession. In Reinharts estimation
(1984: 801), the two are identical (see also Dry 1981: 19), and she tries to support her equating of the concepts by addressing possible objections. The first
objection is raised by examples such as [4.12], in which the backgrounded action expressed by the subordinated clause appears to be temporally successive
with the surrounding clauses.
[4.12] When I strapped on his helmet, Colin ran out to get his bicycle.
In order to account for examples such as [4.12], Reinhart (1984: 79798) distinguishes between content criteria and linguistic, or sentence-level, criteria with respect to foreground/temporal succession: while the subordinated
clause is temporally successive and, therefore, foregrounded in terms of content, linguistically it is backgrounded because it is expressed by a subordinated
clause. However, the definition of temporal succession given above (4.2.1)
makes Reinharts explanation unnecessary. It is clear that the subordinated
clause does not express temporal succession since it does not have the irreversibility property: the order of the subordinated clause with its matrix clause
does not affect the interpretation (i.e., Colin ran out to get his bicycle when I
strapped on his helmet).
287
288
Chapter 4
Table 4.3. Features of the Saliency Continuum
(see Hopper 1979: 129; Hopper and Thompson 1980: 252)
More Salient
temporal succession
perfective
dynamic
telic
volitional
affirmative
realis
nonanaphoric
identity of subject maintained
human topics
agent high in potency
object totally affected
object highly individuated
Less Salient
temporal overlap
imperfective
nondynamic (descriptive)
durative
nonvolitional
negative
irrealis
anaphoric
frequent change of subject
nonhuman topics
agent low in potency
object not affected
object nonindividuated
to boundedness (e.g., perfectivity, telicity) are generally more salient, saliency is a broader and different sort of concept; thus, it cannot be limited
solely to the semantic parameters that determined (un)boundedness.
4.3. The Semantics of Temporality in BH Discourse
Having examined the semantics of temporal succession and pragmatics of
the foreground-background distinction above (4.2), I turn back to the question of the relationship between these elements of discourse organization and
the wayyiqtol and irrealis qatal. The argument of this section is to some extent
negative, inasmuch as it is framed against and contrasts with the traditional and
discourse approaches to these verbal forms. Because of the foil of these other
approaches, it is also somewhat disjointed. I begin by addressing the role of
wayyiqtol in prose narrative discourse (4.3.1). I follow this up in 4.3.2 with a
complementary analysis of the conjugation in poetry, which helpfully clarifies
both the narrative function of wayyiqtol as well as illustrating the symbiotic
interaction of semantic meaning and discourse function (see quotation from
Fleischman, above, 4.1). The following section (4.3.3) turns to the question
of irrealis qatal and its discourse functions. Here, I especially use Longacres
careful analysis of the form to illustrate the shortcomings of discourse approaches to the form. Finally, I briefly examine the bifurcated functions of
the 3ms forms of the copular verb in wayyiqtol and irrealis qatal (4.3.4),
which conveniently illustrate the distinction between semantics and discourse
pragmatics. All this is a sort of clearing of the decks and sorting out of
preliminary issues in preparation for 4.4, in which I present an alternative,
semantics-based analysis of temporality of discourse.
289
And she rose and her daughters-in-law and she returned from
the steppes of Moab, for she had heard in the steppe of Moab
that Yhwh had visited his people by giving bread to them. And
7. Other examples include Josh 2:23 (the two men returned and came down from the
hill country . . .) and Deut 8:3 (he afflicted you and made you hunger). The example in
2Sam 11:17 cited in Cook (2004b: 259) is contestable: the sequence of fought . . . fell ...
died could be understood as successive, since each precludes the continuation of the previous situation; however, the plural subjects of the first two events suggest a degree of overlap.
290
Chapter 4
she went out from the place where she had been and her two
daughters-in-law with her and they went on the way to return
to the land of Judah.
Implicitly bound activities allow for a more general type of temporal
overlay that might be termed coincidental, as illustrated in example [4.15]
(see also Josh 8:4; 18:8), where Rahabs lowering of the men through her
window and her directions to them cannot be temporally interpreted in that
order; rather, a reasonable understanding would be that the complex activity of
lowering them (i.e., getting the rope, tying it off, etc.) overlaps with her giving
them directions.
[4.15] Josh 2:1516
And she lowered them with a rope through the window, for
her house was in the city wall and she was living in the wall;
and she said to them Go to the hill-country lest the pursuers
overtake you.
Stative predicates cannot be bound by wayyiqtol, though frequently they are
interpreted as having an initial temporal bound (i.e., inchoative state become).
However, in example [4.16], real-world knowledge tells us that Jacob did not
fall in love (inchoative) during the course of his coversation with Laban (see
Hatav 1989: 496); nor is it appropriate to render the wayyiqtol form as past perfect (e.g., Now Jacob had fallen in love with Rachel; so the reb, nab). Rather,
in the course of his month-long stay with Laban, Jacob fell in love and now,
where the fact is important to the story, it is reported that Jacob loved Rachel
(so the njps, nrsv).
[4.16] Gen 29:14b18
He stayed with him a month. And Laben said to Jacob, Because you are my kin, should you therefore serve me without
compensation? Tell me what your wage should be? Now
Laben had two daughters. The name of the older was Leah and
the name of the younger was Rachel. And the eyes of Leah
were soft/weak, but Rachel was lovely and beautiful. Now
Jacob loved Rachel, and he said, I will serve you seven years
for Rachel, your younger daughter.
291
In example [4.17], Esaus eating and drinking are simultaneous or alternating. This simultaneous overlap contrasts with the strict temporal succession
expressed by the second two wayyiqtolsand he rose and he left.
[4.17] Gen 25:34
Jacob gave to Esau bread and lentil stew, and he ate and he
drank and he rose and he left, and Esau despised his birthright.
Finally, a different sort of overlap exists in the case of verbal hendiadys, in
which the verbs refer to the same event. This type of overlap is found especially with verbs of speaking as in example [4.18] (see Miller 2003: 14757
for other examples involving reported speech).
[4.18] Exod 6:2
God spoke to Moses and said to him, I am Yhwh.
13
So Joash lay down (wayy.3ms) with his fathers, and Jeroboam sat (qtl.3ms) upon his throne; and Joash was buried
(wayy.3ms) in Samaria with the kings of Israel. 14Now Elisha
had become sick (qtl.3ms) with the illness of which he would
die, and King Joash of Israel had gone/went down (wayy.3ms)
8. For other examples, see Josh 13:833; 2 Kgs 7:67, 1520; and Baker 1973: 2353.
292
Chapter 4
to him, and (had) wept (wayy.3ms) before him, and (had) said
(wayy.3ms), My father, my father! The chariots of Israel and
its horsemen! 15And Elisha (had) said (wayy.3ms) to him,
Take (impv.ms) a bow and arrows; so he had taken/took
(wayy.3ms) a bow and arrows. 16And he (had) said (wayy.3ms)
to the king of Israel, Grasp (impv.ms) the bow; and he (had)
grasped (wayy.3ms) it. And Elisha (had) laid (wayy.3ms) his
hands on the kings hands. 17And he (had) said (wayy.3ms),
Open (impv.ms) the window eastward; and he (had) opened
(wayy.3ms) (it). And Elisha (had) said (wayy.3ms), Shoot
(ms.impv); and he (had) shot (wayy.3ms). And he (had) said
(wayy.3ms), The Lords arrow of victory and the arrow of victory over Aram! For you should fight (qtl.2ms) the Arameans
in Aphek until (you) make an end (inf) (of them). 18And he
(had) said (wayy.3ms), Take (impv.ms) the arrows; and he had
taken/took (wayy.3ms) (them). And he (had) said (wayy.3ms) to
the king of Israel, Strike (impv.ms) the ground; and he (had)
struck (wayy.3ms) three times, and (had) stopped (wayy.3ms).
19
And the man of God had become/became angry (wayy.3ms)
with him, and (had) said (wayy.3ms), (You should have) struck
(inf) five or six times; then you would have struck (qtl.2ms)
Aram until (you) made an end (inf) (of it), but now you will
strike (yqtl.2ms) Aram only three times. 20And Elisha (had)
died (wayy.3ms), and they (had) buried him (wayy.3ms).
Notice, however, that in English example [4.13] above, the past-in-the-past
temporal deixis carries through the passage even with the alternation between
Past Perfect and Simple Past verbs. The case in BH is similar: the initial past
perfect is signaled semantically by the past-perfect qatal ( he had become
sick), the meaning of which is clear from the discourse, which moves from
reporting Joashs death to narrating events prior to Joashs death; the wayyiqtol
verbs form a subsequent storyline through which the past-in-the-past temporal
frame is maintained (see 3.4.1 and C.S. Smith 2003: 94).
There are other cases in which a flashback storyline expressed by wayyiqtols is signaled by a temporal protasis instead of a past perfect qatal (e.g., Gen
19:2829; 2 Sam 4:4; 1 Kgs 11:1522). The exigencies of linearly recounting
two parallel storylines (such as overlapping reigns of kings in 2 Kgs 14:116)
also implicates a past perfect sense for wayyiqtol forms. Both these cases are
illustrated by example [4.20] (see Talmon 1978).
293
Story line 1: Many days passed (wayy.3ms), and the word
of Yhwh came (qtl.3ms) to Elijah, in the third year (of the
drought), saying, Go (impv.ms), present yourself (impv.ms) to
Ahab, and I will send (yqtl.1s) rain on face of the land. 2And
Elijah went (wayy.3ms) to present himself (inf) to Ahab.
Story line 2: Now the famine was severe (qtl.3ms) in Samaria.
3
And Ahab summoned (wayy.3ms) Obadiah, who was in charge
of the palace.
Background: Now Obadiah greatly feared (qtl.3ms) Yhwh,
4
and (
) 9 when Jezebel was killing (inf) the prophets of
Yhwh, Obadiah took (wayy.3ms) a hundred prophets and hid
them (wayy.3ms), fifty men in (each) cave, and sustained them
(qtl.3ms) with bread and water.
Story line 2 (contd.): 5And Ahab said (wayy.3ms) to Obadiah,
Go (impv.ms) through the land to all the springs of water and
to all the wadis; perhaps we will find (yqtl.1p) grass and keep
the horses and mules alive (qtl.1p), and we will not have to
destroy (yqtl.1p) some of the animals.
294
Chapter 4
295
In the third month of the exodus of Israel from the land of
Egypt, on the very day, they came (qtl.3p) to the wilderness of Sinai. They set out (wayy.3mp) from Raphidim and
came (wayy.3mp) to the wilderness of Sinai and encamped
(wayy.3mp) in the wilderness; Israel encamped (wayy.3ms)
there before the mountain.
The correlation between wayyiqtol and the foreground fairs much better than
the case for temporal succession discussed above. Many languages in the
world feature a narrative verb or tense, which is employed in narrative discourse to express the basic story line or foreground (see Longacre 1990: 109).
Within BH narrative discourse, there is overwhelming data to support the assertion that wayyiqtol is a narrative verb in this sense. In other words, wayyiqtol
is the default verb in narrative discourse, and as such, presents (i.e., marks)
foregrounded events.12
Recognition of wayyiqtol as discourse-pragmatically marking the foreground in narrative is exegetically significant when it occurs in contexts where
we would more naturally expect a background construction to be employed, as
in the account of Jacobs marriage to Labans daughters, the relevant verses of
which are given in example [4.23].
12. Heimerdinger (1999: 22325) is the only scholar I have found who takes issue with
this view that wayyiqtol foregrounds events in Biblical Hebrew narrative. His dissension,
however, is based on a different understanding of foregrounding from the one proposed here.
Heimerdinger defines foreground based on Grices maxims of conversation (i.e., quantity,
quality, relation, and manner; see Levinson 1983: 97166): the speaker or writer throws
certain events into relief when their inclusion in the discourse violates one of the maxims of
conversation. In Heimerdingers words (1999: 240), Foregrounding in Old Hebrew narratives could be described as special salience. In the development of a story, there are special
events which are more arresting than others, because they happen out-of-the-blue. By
contrast, I am understanding the foreground as the default (or unmarked) mode of recounting events in narrative, often marked (as in BH) by means of a dominant narrative verb;
background is marked by departures from the default mode of narration (see 4.2.2 above).
Within the dominant foreground, events may nevertheless be determined to be more or less
important to the development of the theme, as Heimerdinger observes in his criticism of
Longacre (1999: 7680); but this is a different (literary) matter.
296
Chapter 4
And Jacob did so and he fulfilled this seven years and he
gave to him Rachel, his daughter, for a wife. And Laban
gave (wayy.3ms) to Rachel, his daughter, Bilhah his maid
to be a maid for her. And he came in to Rachel also, and he
loved Rachel more than Leah.
In each of these excerpts, the use of to report Labans gift of a maidservant
seems out of place; we may justifiably expect a background construction using qatal, such as . Thus, this departure is exegetically significant: by
using wayyiqtol forms, and intentionally interrupting the reports of the two
sisters marriages to Jacob, the narrator highlights and foreshadows the role of
the maidservants as surrogate mothers in the sororial feud between Leah and
Rachel (see Collins 1995: 13233).
The narrator similarly capitalizes on the unexpected use of a foreground
wayyiqtol in the passage cited above, example [4.16], where Jacobs developed
affection for Rachel might have been portrayed with a qatal background construction such as
Now Jacob was in love with Rachel in
this context (cf. Gen 37:3). A construction of this sort, however, would obscure
the central motif in the ensuing narrative of Jacobs love for Rachel.
Although the narrative verb wayyiqtol consistently marks foregrounded
events, it does not automatically follow that every other verb form is limited
to backgrounded events (cf. Longacre 2003: 65). Notably, wayyiqtol and qatal
often appear together expressing simultaneous foregrounded events, as illustrated in example [4.24] (e.g., Gen 33:16; 45:14; Exod 17:10; Judg 1:25; see
Talmon 1978: 12; Gross 1981).13
13. For other examples of qatal expressing foreground events, see Heimerdinger 1999:
9398.
297
298
Chapter 4
I waited (qtl.1s) patiently for Yhwh, and he turned
(wayy.3ms) to me and heard (wayy.3ms) my cry; he brought me
up (wayy.3ms) from the desolate pit, from the muddy mire,
15. I am not strictly distinguishing poetry and verse in my discussion; my main focus is
on wayyiqtol outside prose narrative, and scholars tend to operate with a simple prose-poetry
distinction in mind for biblical literature. The focus on Psalms is convenient inasmuch as it
contains exemplary biblical poetry in verse form.
16. Only 11 poems contain 4 or more wayyiqtol forms, which together account for almost half of the remaining forms in the book (62 of 139 examples): Ps 7:5, 13, 16 [2]; 29:5,
6, 9, 10; 37:36 [2], 40 [2]; 40:2 [2], 3 [2], 4; 44:3 [2], 10, 1921; 50:1, 6, 1618;
64:810 [2]; 69:11 [2], 12 [2], 21 [2], 22; 80:6, 910 [2]; 109:3, 4, 16, 17 [3], 18
[2], 28; 119:26, 52, 55, 5960, 106, 131, 147, 158, 167. Excluded from the statistics are the
8 examples in the prose-narrative titles (i.e., Ps 34:1 [2]; 52:2 [2]; 54:2; 59:1; 60:2 [2]),
17. Excluding the prose-narrative titles that feature multiple wayyiqtol forms (i.e., Ps
34:1; 52:2; and 60:1) and the 5 narrative poems, a series of 2 or more wayyiqtol forms occur
without an intervening predicate only 21 times: Ps 7:16; 29:56; 37:36, 40; 40:23; 55:67;
64:10; 69:1112, 21; 73:1314; 80:10; 94:2223; 97:8; 109:1718; 144:3.
299
In distress you called, and (then) I rescued you.
c. Ps 97:4b
300
Chapter 4
such, they underscore the motivation for the use of the wayyiqtol in this context
as its overt marking of temporal succession.
[4.27] a. Ps 80:9
A vine from Egypt you pulled out; you drove out nations
and planted it.
b. Ps 55:18
Evening and morning and noonday I was lamenting and
moaning and he heard my voice.
c. Ps 144:3
Yhwh, what is man that you made yourself known to him,
the son of man that you considered him?
d. Ps 136:1011
(Praise be) to him who struck Egypt in their firstbornhis
steadfast love is everlastingand brought Israel out from
their midsthis steadfast love is everlasting.
Above (4.2.1), I argued that temporal succession is semantically effected
by boundedness but is the pragmatically default interpretation of narrative discourse. What I propose here is that the association of this pragmatic-based
temporally successive meaning with the narrative wayyiqtol allows the poets
of Psalms to draw on the form to overtly mark temporal succession within a
context dominated by static (temporally speaking), parallelistic poetry. The
contrast between Hebrew narrative and poetry can be well appreciated in light
of Jakobsons (1960: 358) famous definition of poetic function as transforming
sequence into equivalency: equivalence is promoted to the constitutive device
of the sequence. In other words, the relationship between predicates in successive lines (i.e., parallel stichs) in poetry is one of equivalencythat is, they
refer to the self-same eventin contrast to prose narrative, in which successive
predicates refer to successive events. Thus, to force a sequential reading of
events in poetry, the wayyiqtol form is employed with its implicature meaning
of succession from prose narrative.
As I illustrate below (this section), the overt marking of temporal succession is also the main motivation for the use of wayyiqtol in the 5 poems in
which it predominates. But note that the exceptions (i.e., nontemporally successive) among the examples outside these 5 poems, illustrated in example
301
[4.28a], can mostly be explained by other factors that contribute to the verb
choice: in 11 cases, poetic word pairs along with word order and the aesthetics of verb variation dictate the use of wayyiqtol in close succession with
qatal to form a verbal hendiadys or parallel expression (Ps 7:16; 16:9; 29:10;
37:40; 50:1; 73:13; 77:19; 89:20, 39; 90:2; 119:106); similarly, some of the
8 instances of wayyiqtol in simple coordination (i.e., nontemporally successive; Ps 7:5; 29:6; 38:3; 45:8; 80:6; 109:16; 119:55, 167) may be explained
as motivated by complementary word pairings; finally, in 15 cases, wayyiqtol
does not follow anything (in terms of temporal succession) but introduces an
independent past event (8:6; 35:21; 50:6; 64:89; 65:9; 76:3; 81:17; 92:1112;
94:22; 109:1718; 139:11; 148:14). In some cases, such as example [4.28c],
specific discourse concerns may motivate the choice: in Ps 50:6, the wayyiqtol
interrupts Gods speech to make a past report.
[4.28] a. Ps 77:19
The sound of your thunder (was) in the wheel, lightning lit
up the earth, the earth shook and quaked.
b. Ps 45:8a
You (have) loved righteousness and hated wickeness.
c. Ps 50:57
302
Chapter 4
Then the Lord awoke (wayy.3ms), as a sleeper,
as a strong-man shouting (ptc.ms) because of wine.
He routed (wayy.3ms) his enemies;
he placed (qtl.3ms) on them enduring disgrace.
He rejected (wayy.3ms) the tent of Joseph;
he did not choose (qtl.3ms) the tribe of Ephraim.
But he chose (wayy.3ms) the tribe of Judah,
the Mt. Zion, which he loves (qtl.3ms).
Then he built (wayy.3ms) his sanctuary like the heights,
like the earth he founded (qtl.3ms) forever.
b. Ps 106:2830
They attached themselves (wayy.3mp) to Baal Peor,
and ate (wayy.3mp) sacrifices to the dead.
Thus they provoked (wayy.3mp) with their deeds,
and a plague broke out (wayy.3fs) among them.
Then Phineas stood up (wayy.3ms) and interceded
(wayy.3ms)
and the plague was halted (wayy.3fs)
20.The wayyiqtol occurs 185 times in these 5 poems, whereas the qatal and yiqtol each
occur 145 times, and the Participle occurs 66 times.
303
Departures from the constitutive narrative wayyiqtol in these narrative poems are of two basic sorts: in some cases wayyiqtol is avoided at the head of a
line for syntactic (i.e., a negative event) or pragmatic (i.e., focus fronting, new
section head) reasons (e.g., Ps 78:10, 2425, 59; 105:14); or for rhetorical or
poetic reasons events are reported in an order other than the ordo naturalis of
narrative, so the wayyiqtol form is avoided, as illustrated by the examples in
[4.30].
[4.30] a. Ps 105:1618
[ ]
Then he summoned (wayy.3ms) a famine upon the land,
every staff of bread he broke (qtl.3ms).
He had sent (qtl.3ms) a man before them,
Joseph was sold (qtl.3ms) as a slave.
They oppressed (qtl.3ms) his feet with fetters,
his neck entered (qtl.3ms) (a fetter of) iron.
b. Ps 78:4451
304
Chapter 4
305
When you go (inf) today from me, (then) you will meet (irrqtl.2ms) two men near the tomb of Rachel in the territory of
Benjamin in Zelzah, and they will say (irr-qtl.3p) to you, The
donkeys that you went (qtl.2ms) to look for (inf) have been
found (qtl.3p), and behold, your father has abandoned
306
Chapter 4
307
3
If the anointed priest sins (irr-yqtl.3ms), to the guilt of the
people, then he should offer (irr-qtl.3ms) for his sin that he has
committed (qtl.3ms) a bull, son of the herd, without blemish to
Yhwh for a sin offering. 4And he should bring (irr-qtl.3ms) the
bull to the entrance of the tent of meeting before Yhwh, and
he should lay (irr-qtl.3ms) his hand on the head of the bull,
and he should slaughter (irr-qtl.3ms) the bull before Yhwh.
5
And the anointed priest should take (irr-qtl.3ms) some of
the blood of the bull and bring (irr-qtl.3ms) it into the tent of
meeting. 6And the priest should dip (irr-qtl.3ms) his finger in
the blood and sprinkle (irr-qtl.3ms) some of the blood seven
times before Yhwh in front of the curtain of the sanctuary.
7
And the priest should place (irr-qtl.3ms) some of the blood
on the horns of the altar of fragrant incense, which is in the tent
of meeting before Yhwh, but all (the rest) of the blood of the
bull he should pour (yqtl.3ms) at the base of the altar of burnt
offering, which is at the entrance of the tent of meeting.
By contrast, instructional discourse is introduced by an imperative, as in
example [4.34]. In addition, the step-by-step characteristic is less evident
than in procedural discourse: in some cases, there is no logical reason for one
instruction to precede the other other (e.g., there is no reason why the instructions for the ark must precede those for the altar).
[4.34] Exod 25:13, 811; irrealis qatal in instructional discourse
. . .
308
Chapter 4
1Then the Lord spoke to Moses: 2Instruct (impv.ms) the children of Israel that they should take (irr.yqtl.3p) an offering for
me; from everyone whose heart prompts (yqtl.3ms) them, you
should receive (irr-yqtl.2mp) my offering. 3And this is the offering that you should receive (irr-yqtl.2mp) from them: gold,
silver, and bronze. . . . 8And they should make (irr-qtl.3p)
a sanctuary so that I may dwell (irr-qtl.1s) in your midst.
9
According to all that I am showing you (ptc.ms)the pattern
of the sanctuary and all its furniturethus you should make
(irr-yqtl.2mp) it. 10They should make (irr-qtl.3p) an ark of
acacia wood; and its length (should be) two and a half cubits,
its width a cubit and a half, and its height a cubit and a half.
11
You should overlay (irr-qtl.2ms) it with pure gold; inside and
outside you should overlay (yqtl.2ms) it, and you should make
(irr-qtl.2ms) a molding of gold upon it all around.
Despite the high degree of correlation in these passages between irrealis
qatal and foregrounded, iconically ordered procedural/instructional steps, the
interaction of verb forms is too varied to allow a simplistic identification of
irrealis qatal as a temporally successive, foregrounding verb. Other predications appear on the instructional/procedural foreground (e.g.,
he should
pour out [irr-yqtl.3mp] speak [impv.ms] in Exod 25:1) and irrealis qatal
functions in the background (e.g.,
that I may dwell [irr-qtl.1cs] in
Exod 25:8; so Longacre 1994: 55). In addition, larger syntactic structures such
as protasis-apodosis constructions are missed by an analysis of irrealis qatal as
simply a sequential foregrounding verb in non-narrative discourse. Finally, unlike wayyiqtol, which is associated particularly with narrative discourse, nonnarrative is too broad and varied to claim as the proper domain of irrealis qatal.
More-modest claims might be the following: first, irrealis qatal is not native
to narrative discourse as wayyiqtol is, though it often occurs in habitual or subordinate modal expressions within prose narrative; rather, it appears primarily
in reported speech because its irrealis mood is incompatible with realis past
narrative and more appropriate to various statements about alternative events;
second, it is found particularly in series of irrealis events that are iconically
ordered, because of its bounding capabilities associated with being a perfective
gram; third, variation between it and other irrealis forms may in part be attributed to its verb-subject word order and the associated conjunction that appears
with most examples of the conjugation. However, this sort of characterization
provides minimal explanatory help in the analysis of passages with the form
in the Hebrew Bible. In 4.4, I propose a semantic approach to the issue of the
temporal shaping of discourse.
309
4.3.4.Discourse
and
In light of the above treatments of wayyiqtol and irrealis qatal, a brief statement is required with regard to the special cases of
and ( i.e., the 3ms
of these two conjugations). A number of features shared by these forms make
them a special case, foremost of which is that being copular verbs puts them
in a special category beside other predications (so Longacre 2003: 64). The
narrative form
has especially been the focus of numerous studies (see the
survey of previous studies in Harmelink 2004). Here, however, I am narrowly
focused on qualifying the status of
and with respect to the previous
discussions of wayyiqtol and irrealis qatal (4.3.14.3.3) and, at the same
time, clarifying their temporal role in anticipation of the following examination of various discourse specimens, in 4.4.
Hebrew has a zero-copular strategy, and BH evidences at least an incipient
present-parameter pattern of copular employment: in present temporal expressions, the copula is covert or null; in past or future expressions, it is overt
(see 3.2.3.3). Given these analyses, a central function of the copula in BH is
simply to provide a landing site for TAM marking: the copula appears in the
wayyiqtol and qatal conjugations indicating past tense, in yiqtol expressing
future tense, and in irrealis qatal expressing irrealis mood, as illustrated by
examples [4.35ad].
[4.35] a. 1 Sam 1:1
There was a certain man from Ramathaim, a Zophite from
the hill country of Ephraim.
b. Gen 3:1
The serpent was craftier than any animal of the field.
c. Gen 16:12
He will be a wild ass of a man.
d. Gen 2:10
A river (was) flowing out from Eden to water the garden,
and from there it would divide and become four branches.
The in [4.35d] is probably best understood habitually, as translated. However, the semantic blurring between irrealis mood and future tense accounts for
the ambiguity in other instances of , such as in example [4.36].
310
Chapter 4
I will pass through all of your flock today removing from there
every speckled and spotted sheep and every black sheep among
the lambs and every spotted and speckled (one) among the
goats; (that) will be my pay.
More importantly, the
and forms exhibit two distinct functions: a
copular predicate at the clause level, as illustrated above in examples [4.35]
[4.36]; and a discourse-level TAM signal, often at the opening or closing of a
scene or episode (see especially van der Merwe 1999; Joosten [2009] refers
to it as narrative wayhi). The distinction between the clause-level and the
discourse-level /
functions is found in the absence of any arguments
with the latter, as illustrated by example [4.37]:23 in [4.37a], the first
functions to mark the discourse time in a new episode (note the scene shift here
from vv.13, which also begins with
in v.1), whereas the second is a full
copular form with an explicit subject ()
and complement (
).
In example [4.37b], the first-occurring functions as a copula with an expressed subject ( )and complement (),
whereas the second
governs nothing at the clause level but reasserts the habitual sense of the passage at the discourse level (thus avoiding the erroneous interpretation of the
death as describing a specific judges demise).24
[4.37] a. 2 Sam 7:4
[past] During that night, the word of Yhwh came to
Nathan.
b. Judg 2:1819
. . .
23. By making the distinction in this way, my argument is somewhat distinct from and
simpler than van der Merwes (1999: 11314); however, I am in basic agreement with him
regarding the temporal functions of the form, which he describes in the following terms:
anchors a state of affairs to the time-line (i.e., story-line) of a narrative . . . updating the
current reference time of a subsequent scene.
24. The analysis of these forms in the Accordance syntax (currently includes the books
of Genesis, Joshua, Judges, Joel, Amos, Obadiah, Jonah, Nahum, Habakkuk, Psalms, Proverbs, Ruth, Ecclesiastes, Esther, Daniel, and Ezra) shows that this discourse
occurs about
one and one quarter times more frequently than the copular form with complement and that
discourse occurs about one and one half times more often than its copular counterpart.
311
[past] After these things, it was told to Abraham: Behold,
Milcah also has borne children to Nahor, your brother.
c. Isa 7:20
On that day, the Lord will shave with a razor hired beyond
the river, the king of Assyra, the head and the hair of the
feet and also the beard it will remove.
312
Chapter 4
d. Isa 7:21
[hab] On that day, a man will preserve a young cow and
two sheep.
Minimal pairs such as these, few though they be, not only support the syntactical analysis of the prepositional phrases in examples [4.37ab] but also support
the need to identify word order in terms of subject and predicate/verb and not
simply clause-initial versus non-clause-initial.
Longacre (2003: 67-68) argues that never presents foregrounded material, despite being the narrative verb. This sort of view requires revision based
on the analysis presented here of a bifurcated
: the question of foreground
is moot for discourse-functioning
, since it simply (re)asserts the temporal
reference time; but as a copula, despite the stativity of ( see Gen 29:18 cited
in example [4.16] above), the narrative wayyiqtol form foregrounds the predicate, contrasting with the use of in qatal to express past, backgrounded
events (frequently in negative or subordinate clauses). This foreground-background contrast is illustrated by examples [4.39ab]: the
forms in [4.39a]
express salient events that move the storyline forward, evident especially from
the ingressive interpretation in one case and also from the fact that his master observes Josephs condition in the following verse (v.3); in [4.39b], the
background status of the SV clauses is evident from their contrast with the
neighboring verses (Gen 7:5, 7), both of which contain wayyiqtol predicates.
[4.39] a. Gen 39:2
Yhwh was with Joseph, and he became a successful man
and stayed in the house of his master, the Egyptian.
b. Gen 7:6
Now Noah was six hundred years old, and the flood became water upon the earth.
The case of , as with irrealis qatal, is almost the reverse: while the
foreground-background distinction does not apply to the discourse-functioning
, the copular is almost entirely restricted to background expressions,
no doubt in large measure due to the association of its irrealis meanings with
subordinate types of modality.
4.4. The Temporal and Modal Interpretation of Discourse
Up to this point, the discussion in this chapter has largely been deconstructive: in 4.14.2, I made the case that a discourse approach devoid of a
313
314
Chapter 4
:
They said, All that Yhwh has said, we will do; that is, we
will obey.
d. 1 Kgs 18:41; conjunctive waw
Go, eat and drink.
Because the so-called consecutive/conversive waw appears only immediately before a finite verb (see 2.1 for an overview), word order is viewed as
integrally related to the multivalency of the waw conjunction.28 For example,
GKC (Kautzsch 1910: 142a) states that, [i]n the great majority of instances,
however, the position of the subject at the beginning of a verbal-clause is to be
explained from the fact that the clause is not intended to introduce a new fact
carrying on the narrative, but rather to describe a state. Waltke and OConnor
(1990: 650) describe word order as the major device for distinguishing between the conjunctive (including the conjunctive-sequential) and disjunctive
waws: namely, the conjunctive and conjunctive-sequential waws stand immediately before a verb, whereas the disjunctive waw stands immediately before
some word other than a verb.
This standard treatment is problematic and unsatisfactory. It is problematic
because it places too much semantic weight on the waw conjunction and word
order. Steiner (2000) has asked whether the waw conjunction has many meanings, one meaning, or no meaning at all. I incline toward the latter view with
respect to verbal coordination: the waw conjoins verbal clauses at the same
syntactic level but contributes no real semantic content to the coordinative
construction (as illustrated by the absence of and in many of the gloss translations in the book). The difficulties inherent with placing too much semantic
weight on word order is illustrated by Peckhams article on the Hebrew verbal
system in which he made word order a central component for distinguishing
not only verbal sequence but tense, aspect, and mood in general (Peckham
1997; see 2.4.2.2). This is not to say that word order cannot or does not serve
to signal semantic distinctions; my own argument for a SV: VS distinction
of realis: irrealis mood demonstrates what sort of role word order might play
(see 3.3).
The traditional treatments of interclausal relationships are likewise unsatisfactory, because their accompanying taxonomies of examples illustrate but
fail to explain interclausal relationships in Biblical Hebrewa weakness that
28. Joon (2006: 115a) states that waw is multivalent only when attached to finite verbal forms.
315
they share with the discourse approaches discussed above (4.1). The reason
that they cannot explain the inner logic of Biblical Hebrew interclausal relationships is precisely because they have focused too exclusively on the waw
conjunction and word order as determinative of the relationships. In keeping
with my prioritization of semantics versus discourse, here I approach the issue
from the other direction, by beginning with a taxonomy of temporal relationships (semantics) that may be expressed among successive clauses (discourse).
This approach is in keeping with my dismissal of the whole notion of a multivalent waw and the accompanying association of waw with either TAM distinctions (conversive waw) or succession (consecutive waw). Although some
coordinative functions included in the traditional discussions of interclausal
relationships are more than temporal (e.g., epexegetical), they are never less
than temporal. That is, temporality can always be measured between successive clauses, and to describe the temporal relationships is to address almost all
of the traditional categories of interclausal relationships (at the very least, for
prose narrative and narrative poetry).
A useful starting point for developing a semantics-based theory is to study
Discourse Representation Theory (DRT; see Kamp and Ryle 1993; Kamp,
Genabith, and Ryle 2011; see also 1.5.1 above) as having developed from an
interest in the semantic features of the interclausal relationships that traditional
semantic theories were unable to treat adequatelynamely, anaphor and temporal succession (i.e., movement of the reference time). Two insights related to
DRT are particularly relevant here. The first, which lies at the center of DRT as
well as other dynamic theories of meaning, is that events need to be interpreted
in context and that events and their contexts have a reciprocal relationship. As
recently stated by Kamp, Genabith, and Reyle (2011: 125),
the interaction between context and utterance is reciprocal. Each utterance contributes (via the interpretation which it is given) to the context in which it is
made. It modifies the context into a new context, in which this contribution is reflected; and it is this new context which then informs the interpretation of whatever utterance comes next.
The second insight was the reinterpretation of Reichenbachs idea of reference time in ways that demonstrate the complex temporal information contained in TAM grams (e.g., Declercks [1986] time of orientation and time
referred to; see 1.2.3.4), which pointed the way toward relating events in
discourse in terms of reference time (see 1.5).
C.S. Smiths (2003) theory of discourse modes is rooted in DRT and especially focused on the temporal relationships among successive clauses.
Thus, her treatment is a useful starting point for an analysis of interclausal
temporal relationships in BH. In contrast to Longacres categories, which are
316
Chapter 4
insufficiently different from genre distinctions (so OConnor personal communication, 2006), Smith focuses on the temporal contours of smaller (local)
stretches of discourse, positing a nonexhaustive list of five discourse modes,
given in table 4.4. These five types are distinguished by situation type, temporal deixis, and temporal progression.
Smiths distinction of situation types differs from the standard divisions
(state, activities, accomplishment, and achievement). Rather, she notes that the
pertinent distinguishing feature in her discourse modes is between eventualities and general statives: the former includes both dynamic and stative predicates, but the statives refer to specific or particular states; by contrast, general
statives are habitual and generic expressions, which refer to kinds rather than
particulars (see Cook 2005). The latter (general statives) can also include both
dynamic and stative predicates.
The three parameters employed by C. S. Smith to distinguish discourse
modesnamely, situation type, temporal deixis or anchoring, and temporal
progressionlead to Smiths (2003: 9397) identification of three temporal
relationships that may be expressed among successive clauses: continuity, deixis, and anaphor, which are illustrated with her examples in [4.41][4.43].
Example [4.41] illustrates continuity, in which the events are interpreted as
temporally successive with one another. All of the events reported in this brief
narrative excerpt (each indicated by a parenthetic, numbered e) are interpreted
as happening in succession, one after another. Note, however, that the one stative situation (in contrast to the other, dynamic events) in the passage, which
is indicated by the parenthetic s1, temporally overlaps with the surrounding
events in the narrative.
[4.41] Continuity (narrative)
She put on her apron (e1), took a lump of clay from the bin
(e2), and weighed enough for a small vase (e3). The clay was
wet (s1). Frowning, she cut the lump in half with a cheese-wire
to check for air bubbles (e4), then slammed the pieces together
much harder than usual (e5). A fleck of clay spun off (e6) and
hit her forehead, just above her right eye (e7).
In passages with deictic temporality, illustrated by example [4.42], events
are interpreted as temporally located with respect to some fixed point such as
the narrators or, more often, the speakers deictic center. In the example from
C.S. Smith in [4.42], the temporal interpretations of the various events are
separately related to the speakers deictic center rather than in any way being
temporally located relative to one another.
317
Situations
eventualities,
specific states
eventualities,
general states
Temporality
temporally located,
dynamic
Progression
in narrative time
anchored to speech time
spatial: scene or object
abstract entities,
general states
atemporal
metaphoric
318
Chapter 4
319
modal succession. Temporal continuity or succession is the default interpretation of narrative, as illustrated in example [4.44], in which the bracketed event
[e0<CN] denotes some past deixis anchoring, prior to the discourse fragment
(i.e., some event e0 precedes the narrators deictic center CN).
[4.44] Gen 21:19; temporal continuity
God opened (wayy.3ms; [e0 < CN], e0 < e1) her eyes and she
saw (wayy.3fs; e1 < e2) a well of water and she went (wayy.3fs;
e2 < e3) and filled (wayy.3fs; e3 < e4) the skin of water and she
gave the lad a drink (wayy.3fs; e4 < e5).
Modal continuity is the dominant interpretation in irrealis mood procedural
materials, in which an agent is obligated to carry out actions in a particular
sequence (e.g., tabernacle construction in Exodus 2531, sacrificial directions
in Leviticus 16). An example is [4.45], which begins with deixis anchoring
of the first event in the series to the speakers deictic center (God or Moses).
Throughout the passage, each event succeeds the previous one, often logically
dependent on the previous events completion. The only exception to this pattern is the overlap between e2 and e3 (that is,
and in v.11): no
specific order is given for overlaying the inside and outside of the ark; this temporal overlap is conveyed by the combination of inverted word order, the concomitant switch to irrealis yiqtol for e3, and the repetition of the verbal lexeme
. In other words, the departure from the pattern of continuity is effected by
a combination of semantic, syntactic, and discourse-pragmatic factors.
320
Chapter 4
They shall make (qtl.3p; ROCNe1) an ark of acacia woodtwo
and a half cubits its length, a cubit and a half its breadth, and
a cubit and a half its height. And you shall overlay (qtl.2ms;
ROe1 < e2) it with pure gold on its inside, and on its outside you
shall overlay it (yqtl.2ms; ROe3e2). Then you shall make
(qtl.2ms; ROe3 < e4) upon it a gold molding round about. And
you shall cast (qtl.2ms; ROe4 < e5) for it four gold rings and
place (qtl.2ms; ROe5 < e6) (them) upon its four feettwo rings
(shall be) on one of its sides and two (shall be) on the other of
its sides. Then you shall make (qtl.2ms; ROe6 < e7) poles of
acacia wood and overlay (qtl.2ms; ROe7 < e8) them with gold
and insert (ROe8 < e9) the poles into the rings on the sides of
the ark, to lift the ark by them.
Procedural legal code is frequently casuistic; that is, it is introduced with a
conditional structure, as illustrated by the excerpt from Leviticus 1 in example
[4.46]. In this case, the first three predicates are not successive: the protasis
e1 is null copular, and thus stative and overlapping with what follows; the
two yiqtol forms that begin the apodosis provide further guidelines for the offeringwhat and whereand a continuous interpretation is avoided by the
topicalizing word order and the repetition of the lexeme and object (
and he must offer it). The switch from irrealis yiqtol to irrealis
qatal in v.4 begins the procedural portion proper of the lengthy apodosis, in
which the actions must be carried out in the prescribed order. Similar to the
case in example [4.45], in Lev 1:9 in [4.46] the shift to an inverted word order
with irrealis yiqtol yields a non-successive contrastive statement with the
previous clause: while the priests must arrange some pieces on the altar, they
must wash other specific parts of the animal before burning the entire animal
as a burnt offering.
321
If his offering (is) (null; RCCNe1) a burnt offering from the
herd, a male without blemish he must offer it (yqtl.3ms; ROe1
e2); at the entrance of the tent of meeting he must offer
(yqtl.3ms; ROe2 = e3) it for his acceptance before Yhwh. He
should lay (irr-qtl.3ms; ROe3 < e4) his hand upon the head of
the burnt offering, and it will be accepted (irr-qtl.3ms; ROe4<
e5) on his behalf as atonement for him. Then he shall slaughter
(irr-qtl.3ms; ROe5 < e6) the bull before Yhwh, and Aarons
sons the priests shall offer (irr-qtl.3p; ROe6< e7) the blood and
shall dash (irr-qtl.3ms; ROe7 < e8) the blood round about upon
the altar that is at the entrance of the tent of meeting. Then he
shall flay (irr-qtl.3ms; ROe8<e9) the burnt offering and cut it
up (irr-qtl.3ms; ROe9 < e10) into its pieces. Then the sons of
Aaron the priest shall place (irr-qtl.3p; ROe10 < e11) fire upon
the altar and arrange (irr-qtl.3p; RO e11 < e12) wood on the fire.
Then Aarons sons the priests shall arrange (irr-qtl.3p; ROe12<
e13) the pieces with the head and the suet upon the wood that is
on the fire on the altar; but its entrails and its legs he shall wash
(irr-qtl.3ms; ROe13 e14) with water. Then the priest shall turn
the whole into smoke (irr-qtl.3ms; ROe14 < e15) on the altar as
a burnt offering, an offering by fire of pleasing odor to Yhwh.
It is apparent from both of these examples that a simple correlation cannot be
made between verb conjugation and temporal or modal interpretations. Rather,
within a given discourse context, the two irrealis forms contrast in pragmatic
ways that are similar to the wayyiqtol-qatal contrast in narrative modes (e.g.,
example [4.24]).
The future sense in these two passages (examples [4.4546]) derives
from the irrealis status of the events: the agents must bring about situations
that are yet to be realized. They are therefore akin to future temporal expressions. However, modal continuity is also present in non-past habitual passages,
which are the irrealis counterpart to realis narrative. In the habitual passage
in example [4.47], the habitual events are temporally anaphorically anchored
322
Chapter 4
by the narrative context in which they appear (i.e., the first irrealis event is
modally accessible from some event prior to it in the past-narrative discourse,
[e0<CN]). After the habitual modal relationship is established by the first irrealis qatal, subsequent irrealis forms express the habitual actions in succession,
each being related to the preceding event.
[4.47] Gen 29:3 (modality continuity in habitual narrative)
All the flocks would gather (irr-qtl.3p; [e0 < CN], RHe0e1)
there, then they would roll (irr-qtl.3p; RHe1 < e2) the stone
from upon the mouth of the well; then they would water (irrqtl.3p; RHe2 < e3) the flock; then they would replace (irrqtl.3p; RHe3 < e4) the stone upon the mouth of the well.
Deixis refers to the anchoring in a personal deictic center. Temporal
deixis is associated particularly with the beginning of narrative passages, in
which the first event is temporally anchored to the narrators deictic center
(CN), and with reported speech, in which events are predominantly anchored
to the deictic center of the speaker (CS). Both of these types of temporal deixis
are illustrated by example [4.48]: notice that the narrative verbs (e 1 and e5) are
temporally continuous with one another and frame the reported speeches, each
of which receives a temporal deixis interpretation: that is, each of the events
in the reported speeches is temporally and individually related to the speakers
deictic center.
[4.48] Gen 24:2324; temporal deixis in reported speech
He said (wayy.3ms; [e0 < CN], e0 < e1), Whose daughter (are)
(null; CS e2) you? Please tell (impv.fs; ROe3CS) me. Is (exist;
CS e4) there at your fathers house a place for us to spend
the night? She said (wayy.3fs; e1 < e5) to him, I am (null;
CSe6) the daughter of Bethuel, son of Milcah, whom she
bore (qtl.3fs; e6 < CS) to Nahor.
Modal deixis, by analogy with temporal deixis, locates the modal source
as the speakers personal deictic center (CS)whether this is the source of
obligation, wish, volition, or epistemic judgment. Thus, this discourse mode
aligns with epistemic and subjective deontic modality in contrast to dynamic
and objective deontic modality (see Verstraete 2007; see table 3.7, p.255).
As argued in 3.3.2 above, the directive system (Imperative and Jussive) in
323
Yhwh said (wayy.3ms; [e0 < CN], e0 < e1) to Moses, Get up
early (impv.ms; ROe2CS) in the morning and station yourself
(impv.ms; ROe3CS) before Pharaoh, and then say (irr-qtl.2ms;
ROe3 < e4) to him, Thus says (qtl.3ms; e5 CS) Yhwh, the
God of the Hebrews, Dismiss (impv.ms; ROe6CS) my people so
that they may serve me (yqtl.3mp; ROe7CS).
Notice that in this passage the deictic center shifts based on the level of embedded speech or discourse pragmatic considerations. For instance, the direct
commands to Moses in e2 and e3 are anchored to Gods modal deictic center
(followed by the modally continuous irrealis qatal in e4),31 as also is the quotation of Gods command in e6. By contrast, the event in e7 is an expression of
permission anchored to Pharaohs deictic center.32
Anaphor refers to the anchoring (temporal or modal) of an event in some
other event reported earlier in the discourse. In temporal anaphor, the event
is temporally associated with (i.e., included within) a previously established
event time, as in the case of static descriptions. Example [4.50] contains two
examples of this sort. In this excerpt, the initial
is the discourse marker
of past time and not the copular predicate. Note, in support of the discourse
analysis, that if it were treated as a copular, as a stative it would require interpreting the event of the sons names as included within one of the events in the
previous verse (see C.S. Smith 2003: 94)either the appointing of his sons or
Samuels growing old (v.1)neither of which makes sense. By contrast, the
three null-copula events are anaphorically related to the time set by this discourse
e1 (i.e., as unbounded events, they include the reference time of e1).
Thus, the discourse
functions in a way that is similar to example [4.50]:
30. The messenger formula is analyzed here as a performative (the material
equivalence sign indicates that the event time is coincident with the speech); however, a
past or perfect interpretation is equally possible: Thus Yhwh (has) said . . ..
31. While e23 implies the same continuity of modal events as e34, that interpretation
derives from the pragmatics of the iconic ordering of the events rather than the semantic
interpretation signaled by the verbal form.
32. That is, dismissing the people gives them leave or permission to serve Yhwh. Alternatively, this could be analyzed as an (objective) dynamic modal expression (see below on
generic modal): so that they can (might be able to) serve me.
324
Chapter 4
325
The association of generics with the structure of reality is suggested particularly by generics that do not refer to actual events but to potentialities,
such as The bishop moves diagonally, spoken in reference to the game of
chess (example from Carlson 1995: 225). In turn, examples of this sort suggest
that, alongside temporal genericity and on analogy with them, a modal generic
mode helpfully distinguishes dynamic and objective deontic modalities from
epistemic and subjective deontic modalities: whereas subjective modality is
anchored by personal deixis (see above), objective modality leaves the source
of modal authority unexpressed or pragmatically implied from the discourse
context. As Verstraete (2007: 18) states it, in objective modality, the speaker
merely describes a modal relation rather than personally committing to it.
In temporal generic statements, the events are temporally unanchored and
are therefore valid at all temporal locations. For this reason, generic expressions can make use of a variety of different verbal forms (see Cook 2005);
this fact in turn demonstrates that verbal conjugations alone do not determine
whether an expression is generic or not. While the verbal forms retain their
usual TAM contrasts (as argued in Cook 2005), the pragmatics of generic statements (i.e., lack of context, kind-referring noun phrases, etc.) forces a generic
interpretation of the events.33 Thus, the yiqtol forms in example [4.52] retain
the sense normally associated with them in deictic reported speech, despite
receiving a generic interpretation.34
[4.52] Prov 10:12 (temporal generic in proverbial sentence
literature)
326
Chapter 4
8:1a
8:1b
8:1c
8:2a
8:2b
8:2c
8:2d
8:3a
8:3b
8:3c
8:3d
327
The temporal relationship between e1 and e2 is mirrored by that of the second and third clauses (e23): the initial narrative (i.e., foregrounded) clause
(e3) is interpreted anaphorically as being included within the reference time of
e2.35 It is important to notice the complex interaction of semantics and context:
the narrative wayyiqtol in e3, although past tense, is not interpreted deictically here (as e1 is), because the context demands an anaphoric interpretation
whereby it is temporally included in the previous event reference time (e2),
which also includes the reference time of e1. Thus, the use of the wayyiqtol is
motivated by its foregrounding characteristic as well as its consonance with
the past temporal context established by the discourse
.
The discourse
in v.2 reestablishes the reference time, which appears
somewhat odd in the middle of an episode (note the pronominal anaphor that
ties the event in this verse to the preceding verse); however, if e 4 were interpreted as a simple copular clause (rather than the discourse
) , it would result
in placing the reference time of the event within the temporal expression e 2,
which does not make sense (Samuels sons had names while he was old). As in
the case of e1, the discourse
in e4 indicates the deictic location of what followsthe pair of verbless clauses in e56, in which it is temporally included.
Here we see at work C.S. Smiths (2003: 94) principle that unbounded events
(e56) are temporally interpreted with respect to a previously established reference point (e4).
The same principle applies to e7, which consists of a null copula and a
Participle (or agentive noun): it includes the reference time of e 4 and, by extension, e1. In turn, whether interpreted as a stative (agentive noun complement:
they were judges) or progressive gram (Participle complement: they were
judging), this event has scope over the remaining events in this descriptive
episode: e7 temporally includes the events of v.3 (e811). These latter events,
which in a different context might be interpreted as temporally successive
events (apart from the negative clause of e8), express the foreground of the
description but are interpreted anaphorically in relation to the stative event e 7.
Episode 2, given in example [4.55], consists of a series of narrative events
temporally interpreted as simple continuity, within which is embedded a number of deictically interpreted speech events. These latter events are quite similar to Smiths report mode, which is dynamic and deictically interpreted
(see C.S. Smith 2003: 96 and example [4.42] above). Given that there is no
35. The idealistic character of her modes is evident in the differences between this passage and her sample description: she notes that, typically, descriptions have a locative
adverbial expression with scope over the whole, and she postulates a tacit durative time
adverbial with similar scope; by contrast, there is no locative expression here (cf. e711,
below) but the overt durative time reference that anchors the anaphorically interpreted
e3 by including the latter within its stative temporal frame.
328
Chapter 4
8:4a
8:4b
8:5a
8:5b
8:5c
8:5d
8:6a
8:6b
8:6c
8:6d
The first three events in this episode (e1214) are therefore typically narrative, being foregrounded, temporally successive events continuously following
the narrative event of episode one (e3): (then) the elders gathered themselves
together; then they came to Samuel; then spoke to him.36 The latter of these
events introduces reported speech consisting of three deictically interpreted
events (e1517). However, each of these three events has a different temporal
structure. The first (e15) is stative; thus it includes the speakers deictic center
within its reference time. By contrast, the following qatal clause (e16) represents a dynamic event, which is given a past temporal interpretation. However, the information from the discourse contextthat the sons are serving
as judges at this time (e7)decides in favor of a perfect interpretation of this
qatal versus a perfective interpretation. Finally, the last event in the reported
speech (e17) is interpreted as modal versus temporal deixis with respect to the
36. C.S. Smith (2003: 94) writes of the narrative mode:
All these situations are temporally located at a time prior to Speech Time, following
narrative convention. Otherwise the information conveyed by tense is simple continuity
[i.e., temporal succession]. The past tense is not interpreted deictically: if it were the
events would be related to Speech Time rather than to each other. Nor do we interpret
the past tense as expressing a series of events successively prior to one another. The continuity function of tense holds for narratives in the present or future as well as the past.
329
speaker; in this case, the speakers deictic center (CS) is the speakers situation
(or world), to which an alternative situation (or possible world) expressed by
the imperative stands in a directive relationship: a speaker-derived obligation
is placed on the addressee to bring about the alternative situation. The same
relation holds for e20 as well.
Event e18 reports Samuels reaction to the speech, thus being successive
with the previous narrative (versus reported speech) clause, e14: the elders
spoke to Samuel, and then the matter was displeasing to him. Although as a
stative, is unbounded, it does not include within its deixis the following
event (e19); instead, the latter appearing in a subordinate clause modifying
, which is anaphorically (i.e., the definite article) linked to the preceding
reported speech (e1417) receives a past perfect interpretation: inasmuch as
they had said. The unbounded character of is evident, instead, from its
relationship with e21, which though successive with the stative e19, does not
wholly follow it but is included within its reference time: what they had said
remained grievous to Samuel when he prayed about it.
The third episode, example [4.56], represents the reverse of the type of
material found in episode two: in place of continuous narrative with embedded
deictic speech, episode three consists of reported speech with a short embedded narrative. The first event in this episode (e22) is successive with the last of
the previous episode (e21): the Lords response to Samuel follows his prayer.
[4.56] Episode 3 (1 Sam 8:79)
e23 (impv.ms) {act, dir}; ROCSe23
e24 (yqtl.3mp) {act, ipfv, decl}; CS e24
e25 (qtl.3p) {ach, pf, neg-decl}; e25 < CS
e26 (qtl.3p) {ach, pf, decl}; e26 < CS
e27 (qtl.3p) {act, pf, decl}; e27 < CS
e28 (wayy.3mp) {ach, pfv, decl}; e28 e27
e29 (wayy.3mp) {act, pfv, decl}; e28 < e29
e30 (ptc.mp) {act, prog, decl}; CS e30
8:7a
8:7b
8:7c
8:7d
8:7e
8:8a
8:8b
8:8c
8:8d
8:9a
8:9b
8:9c
8:9d
330
Chapter 4
in fact, corresponding negative and positive statements. The perfect interpretation of the last (e27) of the three is reinforced by the temporal expression from
the day . . . until this day. This clear indication of a perfect meaning lends
weight to the otherwise somewhat ambiguous qatal forms in the preceding two
clauses: (have) rejected. In turn, the perfect expression in e27 provides the
anaphoric reference time for the following short, two-event narrative sequence,
interpreted as temporal continuity in e2829: the people abandoned God and
then worshiped other deities (though the two events may be thought to occur
relatively simultaneously, they are nevertheless clearly irreversible). Thus this
narrative is embedded within a descriptive passage (e2729), which itself appears within a reportive discourse (e2529).
The following irrealis event (e31), expressed by the directive imperative, is
interpreted as modal deixis, as also is the exceptive clause that follows (e 32).
The final event in this short series of irrealis events, e33, is identical to the
preceding, e32: testifying is further specified in informing them what the
king will do. The final event (e34), in the subordinate clause, has a future yiqtol,
which casts the event after the speakers deictic reference time.
The fourth and final episode in this narrative is the longest (example [4.57]).
It is introduced by the narrative event (e34), which is successive with the previous narrative event of e22. (The narrative events e2829 are excluded from this
sequence of events by their embedding in reported speech.) The progressive
Participle construction in e35 is a relative modifier of the preceding noun, ,
thus containing the event time of e34. The event of e36 forms a hendiadys with
e34, and by its repetition serves to introduce Samuels speech directly.
Samuels predictive speech about the behavior of the king that the people
have requested begins with the deictic orienting yiqtol copula in e37 and a
subordinate relative yiqtol in e38, which is deictically anchored like e37. The
remainder of the description (e3950) is structured with multiple yiqtol forms
of take (e39, e4243, e47) and tithe (e45, e49), each followed by
modifying final clauses, with the exception of the last (e49). The final clauses
consist of irrealis qatals (e4041, e44, e46, e48), except for the subordinate infinitive clauses in v.12, which confirm the final interpretation of the surrounding
irrealis qatals. Notably, all of these sequences of yiqtol-final irrealis qatal,
while successive in themselves, are not successive with each other. Rather,
each yiqtol seems to anchor its own reference time deictically, while the final
irrealis qatal forms are each interpreted as modally anaphoric with respect
to the preceding yiqtols.37 The description ends with a contrastive yiqtol
clause with overt pronoun (e50): But you [in contrast to all that the king will
take from you] will be his slaves.
37. Note that the same formal notation is used for a conditional (example [4.51], p.324)
and a final clause here; in both cases, one event modally follows the anchoring event.
331
8:10a
8:19a
8:19b
8:20a
8:20b
8:20c
8:20d
8:20e
8:21a
8:21b
8:22a
8:22b
8:22c
8:22d
8:22e
8:10b
8:11a
8:11b
8:11c
8:11d
8:11e
8:11f
8:13a
8:14a
8:14b
8:15a
8:15b
8:16a
8:16b
8:17a
8:17b
8:18a
8:18b
8:18c
332
Chapter 4
The warning with which Samuel ends his speech (e5153) consists of a temporal protasis-apodosis construction, each part of which has an overt adverbial
temporal marker on that day that anaphorically links the construction to the stative summary statement in e50: when you cry out on that day,
... Yhwh will not answer you on that day. However, while the protasis is interpreted as modally anaphoric to e50, the apodosis is anchored by the protasis
itself, just like the conditional construction in example [4.51] (p.324), absent
the obligation modal sense. Intervening between these two events is the qatal
relative clause, which is interpreted as future perfect in the predictive context
(e52): whom you will have chosen.38
The response from the people is another reported speech, framed by two
narrative events (e5455) that narratively follow the framing of Samuels speech
(e36). The peoples speech consists of a pattern similar to Samuels: a leading
yiqtol sets the reference time (e56) and is modified by a series of irrealis
qatal final clauses (e5760). Samuel conveys the peoples response to Yhwh,
who responds in turn to Samuel in a series of three narrative clauses (e6163)
that link back as continuous with e55.
Yhwhs brief speech consists of two irrealis clauses: an Imperative (e64) and
an irrealis qatal (e65). The distinction between these two directive expressions
is uncertain. It may be that we should see the second as giving more-precise
instructions regarding the former: listen to their voice; (that is,) you should
make a king for them. This is followed by one final narrative-reported speech
cycle, in which Samuel again addresses the people (e66) and tells them to disband and go home (e67). This final episode nicely illustrates the alternation
of narrative and speech, in which each narrative event links successively to
the preceding, while each reported speech receives a deictically independent
temporal interpretation.
4.4.2. Exodus 12 (Irrealis Instruction)
While irrealis mode interpretations are found scattered throughout the example of narrative discourse above (4.4.1; e.g., conditional constructions,
final clauses, or directive commands), I noted in 4.4 that there are a variety
of modes of interpretation in which irrealis forms predominate, mostly with
a modal sense of obligation.39 Rather than focus here on more-monolithic examples (e.g., modal genericity in Exodus 20, the modal continuity of Exodus
2531 or Leviticus 16), I analyze Exodus 12, which contains a more-complex
series of irrealis clauses, outlining the directions for observing the Passover
and Feast of Unleavened Bread. The episode divisions are based on the logic
38. This overt future perfect interpretation is not required in English: whom you chose
is possible, but within the context in which the people have yet to choose a king, it is implied
that it is past with respect to the future deictic context in which it appears.
39. Habitual stands out as the main exception to this and stands somewhere between
modal continuity and temporal genericity in its character (see Cook 2005: 118 and sources
cited there for discussion of habitual as a subcategory of generics).
333
of the instructions themselves rather than, for example, scene shifts of the sort
used in the analysis of the narrative passage above (4.4.1). The first episode,
Exod 12:15, is given in example [4.58].
[4.58] Step 1: Selection of lamb (Exod 12:15)
e1 (wayy.3ms) {act, pfv, decl}; e1 < CN
e2 (null) {sta, pfv, decl}; CS e2
e3 (null) {sta, pfv, decl}; CS e3
e4 (mp.ipv) {act, deo}; ROCSe4
e5 (yqtl.3mp) {acc, ipfv, deo}; w: ROwe5
e6 (yqtl.3ms) {sta, ipfv, cond}; RCe6e5
e7 (qtl.3ms) {act, pfv, deo}; ROe6 e7
e8 (yqtl.2mp ) {act, ipfv, deo}; ROe8 e5
e9 (yqtl.3ms ) {sta, ipfv, deo}; ROe9 e5
e10 (yqtl.2mp ) {acc, ipfv, deo}; ROe10 e5
12:1
12:2a
12:2b
12:3a
12:3b
12:4a
12:4b
12:4c
12:5a
12:5b
This instructional section regarding the Passover and Festival of Unleavened bread is framed within narrative; hence, the first event (e1) is a narrative
event that temporally locates the event prior to the deictic center of the narrator. Immediately, the account shifts to reported speech, which then carries
through to the end of the account. The verbless clauses of e2e3 are static; the
directive meaning often assigned to them (e.g., this month shall mark for you
. . . njps, nrsv) is a pragmatic implicature (versus a semantic interpretation)
based on the directive context. In turn, these static events set the discoursepragmatic context for the following imperative (e4): (Now) speak . . ..
The following event (e5) introduces another level of reported speech: the
content of what Moses is to say to the people. Notably the verb is prefixed
by the waw conjunction but also preceded by a temporal subordinate phrase,
pointing to the role of conjunction here as indicating irrealis mood VS word order in a manner typical of its customary appearance on irrealis qatal. The morphology allows either a deontic yiqtol or Jussive interpretation: each should
take or let each of them take. One could argue for a Jussive analysis on the
basis that the shift from the Imperative is due the shift to third person, and thus
continuity would dictate remaining with directive forms. However, one might
equally argue that the shift from the Imperative to the irrealis yiqtol is motivated by the shift from subjective modal deixis of Gods command to Moses to
the objective modality of the instructions for observing Passover.40 Following
the latter tack, I have analyzed e5 as having a modal generic interpretation,
40. This uncertainty illustrates the sometimes difficult and always context-sensitive task
of discerning objective versus subjective directive expressions.
334
Chapter 4
though the universal obligation is lexically anchored in the tenth day of the
month.
The remainder of this first episode consist of events tied to the lead instruction in e5. The protasis in e6 is interpreted as modally anaphorically related
to e5, dealing as it does with a special case (family size) regarding the lead
obligation in e5. The apodosis e7 is then anaphorically related to the protasis,
as with other conditional and temporal expressions examined already above.
The choice between irrealis qatal, as here, and irrealis yiqtol for the apodosis
seems to be strictly a word-order-motivated choice: in the yiqtol apodosis in
example [4.51] (p. 324), an adjunct phrase is fronted, whereas here, the irrealis qatal is clause initial. (In neither case is there an overt subject.) The final
three events, e810, are all elaborating the head command in e5, and hence they
are also anaphorically anchored to that event.
In the second episode, in example [4.59], directions are given regarding the
cooking and eating of the pascal lamb. The episode begins with the irrealis
qatal copula ( e11), which is modally continous with the previous event,
e10. This interpretation is reinforced by the temporal phrase
: (So) it must remain in your care (lit., an obligation to you) until the
fourteenth day of this month.
[4.59] Step 2: Cooking and eating a lamb (Exod 12:611)
e11 (qtl.3ms) {sta, pfv, decl}; ROCSe11
e12 (qtl.3p) {acc, pfv, deo}; ROe11 < e12
e13 (qtl.3p) {acc, pfv, deo}; ROe12 < e13
e14 (qtl.3p) {acc, pfv, deo}; ROe13 < e14
e15 (yqtl.3mp) {act, ipfv, decl}; e14 < e15
e16 (qtl.3p) {act, pfv, deo}; ROe15 < e16
e17 (yqtl.3mp) {act, ipfv, deo}; ROe17 = e16
e18 (juss.2mp) {act, neg-dir}; ROCSe18
e19 (yqtl.2mp) {acc, neg-ipfv, deo}; ROe17 < e19
e20 (yqtl.2mp) {act, ipfv, deo}; e20 e19
e21 (yqtl.2mp) {act, ipfv, deo}; ROe21e19
e22 (qtl.2mp) {act, pfv, deo}; ROe22e17
e23 (null) {sta, pfv, decl}; CS e23
12:6a
12:6b
12:7a
12:7b
12:7c
12:8a
12:8b
12:9
12:10a
12:10b
12:11a
12:11b
12:11c
335
Having begun with the static yet continous copular statement in e 11, the episode continues with a series of events interpreted as modal continuity, providing the step-by-step instructions for preparing and eating the lamb (e1214, 16).
The yiqtol relative clause (e15) interrupting this continuity series anaphorically
places the reference time of eating after the application of the blood: in which
you will be eating it. Two deontic events also interrupt the series of modal
continuity: e17 equivalent to e16, providing further details about preparing the
lamb in inverted order with the the former (see example [4.51], p.324, for a
similar structure). The following e18 stands out from the surrounding discourse
as being a negative directive, interpreted as modal deixis. As such, it seems to
function as an apostrophe, a sort of annotation to the directions: make sure not
to eat any of it boiled in water. The following two events, e 1920, then seem to
pick up on the modal continuity mode in unison, as a negative-positive pair of
instructions. The last two verbal clauses in the episode also function as a pair
(e2122), anaphorically related to the e17, providing directions on the manner in
which the meal is to be eaten, and then ending with the null-copula summary
statement, e23.
[4.60] Step 3: Yhwhs Passover (Exod 12:1213)
e24 (qtl.1s) {acc, pfv, cond}; RCe24e17
e25 (qtl.1s) {act, pfv, decl}; Re24 e25
e26 (qtl.1s) {act, ipfv, decl}; e26 e25
e27 (null) {sta, pfv, decl}; CS e27
e28 (qtl.3ms) {sta, pfv, decl}; RCe24 e28
e30 (null) {sta, pfv, decl}; e28 e30
e31 (qtl.1s) {act, pfv, cond}; RCe28 < e31
e32 (qtl.1s) {acc, pfv, decl}; RCe31 < e32
e33 (yqtl.3ms) {acc, ipfv,neg-decl}; CS<e33
12:12a
12:12b
12:12c
12:12d
12:13a
12:13b
12:13c
12:13d
12:13e
336
Chapter 4
12:14a
12:14b
12:14c
12:15a
12:15b
12:15c
12:16a
12:16b
12:16c
12:16d
12:16e
12:17a
12:17b
12:17c
337
12:18
12:19a
12:19b
12:20a
12:20b
4.5.Conclusion
The focus of this chapter is the discourse pragmatics of the BHVS from the
perspective of the semantic analysis in chap.3. In the course of the chapter,
I have been concerned with several issues here: first, that semantics be given
priority over discourse pragmatics in analyzing the BHVS (4.1). The passages
treated in this chapter illustrate the fact that a variety of factors contribute to
the temporal interpretation of the verbal forms and that having the semantics clear first enables the interpreter to distinguish between semantic and
discourse-pragmatic elements.
Second, the negative goal of this chapter is my argument against the traditional point of view that BH has verb forms that explicitly indicate temporal
succession (i.e., wayyiqtol and irrealis qatal; 4.24.3). Again, by beginning
with semantics, I have shown how properly to attribute the interpretation of
temporal succession to semantic and discourse pragmatic factors. In particular,
the two domains interact: on the one hand, the semantics of the past perfective
wayyiqtol makes it the verb of choice in prose narrative, in which discoursebased iconic ordering is the default interpretation; on the other hand, the association of wayyiqtol with prose narrative, which is by default temporally
successive, leads to the implicature of temporal succession in poetic contexts,
where the default mode is static parallelistic rather than successive between
events.
Third, I presented a model with which the temporality of interclausal relationships may be approached from a semantic foundation (4.4). I illustrated
the basic scheme with short excerpts (parade examples) and then follow this
with examinations of lengthier selections of prose narrative (1Samuel 8) and
instructional material (Exodus 12). At this point, the linguistic investigation
fades into philological and literary interpretation. Throughout, my aim has
been to illustrate that a methodology that begins with a rigorous semantic
analysis and then progresses to discourse-pragmatics and finally to general
interpretation (i.e., sensitivity to the literary context) provides a round and
comprehensive interpretation of BH texts that conforms to the principle of
338
Chapter 4
Works Cited
Akman, V.
2006 Situation Semantics. Pp. 98400 in vol. 10 of Encyclopedia of Language and
Linguistics, ed. Keith Brown. Amsterdam: Elsevier.
Albright, W. F.
1937 The Egyptian Correspondence of Abimilki, Prince of Tyre. Journal of Egyptian Archaeology 23: 190203.
1942a A Teacher to a Man of Shechem about 1400 b.c. Bulletin of the American
Schools of Oriental Research 86: 2831.
1942b A Case of Lse-Magest in Pre-Israelite Lachish, with Some Remarks on the
Israelite Conquest. Bulletin of the American Schools of Oriental Research 87:
3238.
1943a Two Little Understood Amarna Letters from the Middle Jordan Valley. Bulletin of the American Schools of Oriental Research 89: 717.
1943b An Archaic Hebrew Proverb in an Amarna Letter from Central Palestine. Bulletin of the American Schools of Oriental Research 89: 2932.
Albright, W. F., and William L. Moran
1948 A Re-interpretation of an Amarna Letter from Byblos. Journal of Cuneiform
Studies 2: 23948.
Allen, Joseph Henry, J. B. Greenough, George Lyman Kittredge, Albert Andrew
Howard, and Benjamin L. DOoge
[1888] 1931 Allen and Greenoughs New Latin Grammar for Schools and Colleges.
Boston: Ginn.
Allen, Robert L.
1966 The Verb System of Present-Day American English. The Hauge: Mouton.
Allwood, Jens, Lars-Gunnar Andersson, and sten Dahl
1977 Logic in Linguistics. Cambridge: Cambridge University Press.
Andersen, Francis I., and A. Dean Forbes
2007 The Participle in Biblical Hebrew and the Overlap of Grammar and Lexicon.
Pp. 185212 in Milk and Honey: Essays on Ancient Israel and the Bible in Appreciation of the Judaic Studies Program at the University of California, San
Diego, ed. Sarah Malena and David Miano. Winona Lake, IN: Eisenbrauns.
Andersen, T. David
2000 The Evolution of the Hebrew Verbal System. Zeitschrift fr Althebristik 13:
166.
Anderson, J. M.
1973 An Essay Concerning Aspect. The Hague: Mouton.
339
340
Works Cited
Andrason, Alex
2010 The Panchronic Yiqtol: Functionally Consistent and Cognitively Plausible.
Journal of Hebrew Scriptures 10. http://www.jhsonline.org/.
2011a Biblical Hebrew Wayyiqtol: A Dynamic Definition. Journal of Hebrew Scriptures 11. http://www.jhsonline.org/.
2011b The Biblical Hebrew Verbal System in Light of Grammaticalization. Hebrew
Studies 52: 1951.
2011c The BH Weqatal: A Homogenous Form with No Haphazard Functions
(Part1). Journal of Northwest Semitic Languages 37/2: 126.
forthcoming The BH Weqatal: A Homogenous Form with No Haphazard Functions
(Part2). Journal of Northwest Semitic Languages. http://www.sun.ac.za/jnsl.
Anstey, Matthew P.
2009 The Biblical Hebrew Qatal Verb: A Functional Discourse Grammar Analysis.
Linguistics 47: 82544.
Anttila, Raimo
1989 Historical and Comparative Linguistics. 2nd rev. ed. Current Issues in Linguistic Theory. Amsterdam: Benjamins.
Arnold, Bill T., and John H. Choi
2003 A Guide to Biblical Hebrew Syntax. New York: Cambridge University Press.
Austin, J. L.
1962 How to Do Things with Words. Oxford: Clarendon.
Auwera, Johan van der, Nina Dobrushina, and Valentin Goussev
2005 Imperative-Hortative Systems. Pp. 29497 in The World Atlas of Language
Structures, ed. Martin Haspelmath and Hans-Jrg Bibiko. Munich: Max
Planck Digital Library.
Baayen, R. Harald
1997 The Pragmatics of the Tenses in Biblical Hebrew. Studies in Language 21:
24585.
Bach, Emmon
1986 The Algebra of Events. Linguistics and Philosophy 9: 516.
Bache, Carl
1982 Aspect and Aktionsart: Towards a Semantic Distinction. Journal of Linguistics 18: 5772.
1985 Verbal Aspect: A General Theory and Its Application to Present-Day English.
Odense University Studies in English. Odense: Odense University Press.
1995 The Study of Tense, Aspect and Action. Frankfurt am Main: Peter Lang.
Baker, David W.
1973 The Consecutive Nonperfective as Pluperfect in the Historical Books of the
Hebrew Old Testament. M.A. thesis, Regent College.
Banfield, A.
1982 Unspeakable Sentences: Narration and Representation in the Language of
Fiction. Boston: Routledge & Kegan Paul.
Barco del Barco, Francisco Javier del
2003 Profeca y Sintaxis: El Uso de las Formas Verbales en los Profetas Menores
Preexlicos (Prophecy and Syntax: The Use of the Verbal Forms in the Pre-
Works Cited
341
342
Works Cited
Bhat, D. N. S.
1999 The Prominence of Tense, Aspect, and Mood. Studies in Language Companion
Series. Amsterdam: Benjamins.
Bickel, Balthasar
2007 Typology in the 21st Century: Major Current Developments. Linguistic Typology 11: 23951.
Bierwisch, M., and R. Schreuder
1992 From Concepts to Lexical Items. Cognition 42: 2360.
Binnick, Robert I.
1991 Time and the Verb: A Guide to Tense and Aspect. Oxford: Oxford University
Press.
Blake, Frank R.
1944a The Hebrew Waw Conversive. Journal of Biblical Literature 63: 27177.
1944b Review of Hebrew Grammar: A New Approach, by Alexander Sperber.
Journal of Biblical Literature 63: 19599.
1946 The Form of Verbs after Waw in Hebrew. Journal of Biblical Literature 65:
5157.
1951 A Resurvey of Hebrew Tenses. Scripta Pontificii Instituti Biblici 103. Rome:
Pontifical Biblical Institute.
Blau, Joshua
1971 Marginalia Semitica I. Israel Oriental Studies 1: 135.
2010 Phonology and Morphology of Biblical Hebrew: An Introduction. Linguistic
Studies in Ancient West Semitic 2. Winona Lake, IN: Eisenbrauns.
Bloch, A.
1963 Zur Nachweisbarkeit einer hebrischen Entsprechung der akkadischen Form
iparras. Zeitschrift der Deutschen Morgenlndischen Gesellschaft 113: 4150.
Bobzin, H.
1973 berlegungen zum althebrischen Tempus System. Die Welt des Orients 7:
14153.
Bodine, Walter R.
1995 Discourse Analysis of Biblical Literature: What It Is and What It Offers.
Pp.118 in Discourse Analysis of Biblical Literature: What It Is and What It
Offers, ed. Walter R. Bodine. Society of Biblical Literature Semeia Studies.
Atlanta: Scholars Press.
Boneh, Nora, and Edit Doron
2008 Habituality and Habitual Aspect. Pp. 32148 in Theoretical and Cross
linguistic Approaches to the Semantics of Aspect, ed. Susan Rothstein. Amsterdam: Benjamins.
2010 Modal and Temporal Aspects of Habituality. Pp. 33863 in Syntax, Lexical
Semantics, and Event Structure, ed. Malka Rappaport Hovav, Edit Doron, and
Ivy Sichel. Oxford: Oxford University Press.
Bordreuil, Pierre, and Dennis Pardee
2009 A Manual of Ugaritic. Linguistic Studies in Ancient West Semitic 3. Winona
Lake, IN: Eisenbrauns.
Works Cited
343
Brockelmann, Carl
190813 Gundriss der Vergleichenden Grammatik der Semitischen Sprachen.
Berlin: Reuther & Richard.
1951 Die Tempora des Semitischen. Zeitschrift fr Phonetik und allgemeine
Sprachwissenschaft 5: 13354.
1956 Hebrische Syntax. Neukirchen-Vluyn: Erziehungsverein.
Brown, Gillian, and George Yule
1983 Discourse Analysis. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press.
Buccellati, Giorgio
1988 The State of the Stative. Pp. 15389 in vol. 58 of Fucus: A Semitic/Afrasian
Gathering in Remembrance of Albert Ehrman, ed. Yol L. Arbeitman. Amsterdam Studies in the Theory and History of Linguistic Science. Amsterdam:
Benjamins.
1996 A Structural Grammar of Old Babylonian. Wiesbaden: Harrassowitz.
1997 Akkadian. Pp. 6999 in The Semitic Languages, ed. Robert Hetzron. London:
Routledge.
2003 Akkadian. Pp. 3942 in vol. 1 of International Encyclopedia of Linguistics,
ed. William J. Frawley. 2nd ed. New York: Oxford University Press.
Bull, William Emerson
1960 Time, Tense, and the Verb: A Study in Theoretical and Applied Linguistics,
with Particular Attention to Spanish. University of California Publications in
Linguistics. Berkeley: University of California Press.
Buth, Randall
1992 The Hebrew Verb in Current Discussions. Journal of Translation and Text
linguistics 5: 91105.
1994 Methodological Collision between Source Criticism and Discourse Analysis:
The Problem of Unmarked Temporal Overlay and the Pluperfect/Non
sequential Wayyiqtol. Pp. 13854 in Biblical Hebrew and Discourse Linguistics, ed. Robert D. Bergen. Dallas: Summer Institute of Linguistics.
Butler, Christopher S.
2003 Structure and Function: A Guide to Three Major Structural-Functional Theories, Part I: Approaches to the Simplex Clause. Studies in Language Companion Series. Amsterdam: Benjamins.
Bybee, Joan L.
1985 Morphology: A Study of the Relation between Meaning and Form. Typological
Studies in Language. Amsterdam: Benjamins.
1998 Irrealis as a Grammatical Category. Anthropological Linguistics 40: 25771.
Bybee, Joan L., and sten Dahl
1989 The Creation of Tense and Aspect Systems in the Languages of the World.
Studies in Language 13: 51103.
Bybee, Joan, Revere Perkins, and William Pagliuca
1994 The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of
the World. Chicago: University of Chicago Press.
344
Works Cited
Works Cited
345
1986 Tense and Time Reference: From Meaning to Interpretation in the Chronological Structure of a Text. Journal of Literary Semantics 15: 1222.
1994 Tense. Pp. 455863 in The Encyclopedia of Language and Linguistics,
ed.R.E. Asher. Oxford: Pergamon.
Cook, John A.
2001 The Hebrew Verb: A Grammaticalization Approach. Zeitschrift fr Alt
hebristik 14: 11743.
2002 The Biblical Hebrew Verbal System: A Grammaticalization Approach. Ph.D.
dissertation, University of Wisconsin.
2003a Review of Focus Structure in Biblical Hebrew, by Katsuomi Shimasaki. Catholic Biblical Quarterly 65: 61819.
2003b Review of the Verbal System in Classical Hebrew in the Joseph Story: An
Approach from Discourse Analysis, by Yoshinobu Endo. Journal of Near
Eastern Studies 62: 6264.
2004a Review of Profeca y Sintaxis: El Uso de las Formas Verbales en los Profetas
Menores Preexlicos [Prophecy and Syntax: The Use of the Verbal Forms
in the Pre-exilic Minor Prophets], by Francisco Javier del Barco del Barco.
Hebrew Studies 45: 33638.
2004b The Semantics of Verbal Pragmatics: Clarifying the Roles of Wayyiqtol and
Weqatal in Biblical Hebrew Prose. Journal of Semitic Studies 49: 24773.
2005 Genericity, Tense, and Verbal Patterns in the Sentence Literature of Proverbs. Pp. 11733 in Seeking Out the Wisdom of the Ancients: Essays Offered to Honor Michael V. Fox on the Occasion of His Sixty-Fifth Birthday,
ed.Ronald L. Troxel, Kelvin G. Friebel, and Dennis R. Magary. Winona Lake,
IN: Eisenbrauns.
2006a The Finite Verbal Forms in Biblical Hebrew Do Express Aspect. Journal of
the Ancient Near Eastern Society 30: 2135.
2006b Review of Biblical Narrative and the Death of the Rhapsode, by Robert S.
Kawashima. Hebrew Studies 47: 43943.
2006c Review of Narrative Structure and Discourse Constellations: An Analysis of
Clause Function in Biblical Hebrew Prose, by Roy L. Heller. Catholic Biblical Quarterly 68: 11617.
2008a The Participle and Stative in Typological Perspective. Journal of Northwest
Semitic Languages 34: 119.
2008b The Vav-Prefixed Verb Forms in Elementary Hebrew Grammar. Journal of
Hebrew Scriptures 8. http://www.jhsonline.org/.
2010 Review of a New Understanding of the Verbal System of Classical Hebrew:
An Attempt to Distinguish between Semantic and Pragmatic Factors, by Rolf
J. Furuli. Journal of Near Eastern Studies 69: 24951.
2012 Detecting Development in Biblical Hebrew Using Diachronic Typology.
Pp.8395 in Diachrony in Biblical Hebrew, ed. Cynthia L. Miller-Naud and
ZionyZevit. Linguistic Studies in Ancient West Semitic 8. Winona Lake, IN:
Eisenbrauns.
forthcoming The Verb in Qoheleth. In Like Goads: Engaging Qoheleth in the 21st
Century, ed. in Mark J. Boda, Tremper Longman III, and Cristian Rata. Winona Lake, IN: Eisenbrauns.
346
Works Cited
Croft, William
2003 Typology and Universals. 2d ed. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press.
Crystal, David
2008 A Dictionary of Linguistics and Phonetics. 6th ed. Oxford: Blackwell.
Curtius, Georg
1846 Die Bildung der Tempora und Modi im Grieschen und Lateinischen sprachvergleichend dargestellt. Berlin: Wilhelm Besser.
1863 Erluterungen zu meiner Griechischen Schulgrammatik. Prague: Tempsky.
1870 Elucidations of the Students Greek Grammar, trans. Evelyn Abbott. London:
John Murray.
Dahl, sten
1985 Tense and Aspect Systems. Oxford: Blackwell.
1994 Aspect. Pp. 24047 in The Encyclopedia of Language and Linguistics,
ed.R.E. Asher. Oxford: Pergamon.
1997 Review of Time in Language, by Wolfgang Klein. Studies in Language 21:
41728.
2000 Tense and Aspect in the Languages of Europe. Empirical Approaches to Language Typology 20-6. Berlin: de Gruyter.
Dallaire, Hlne
2002 The Syntax of Volitives in Northwest Semitic Prose. Ph.D. dissertation, Hebrew Union College.
Davidson, A. B.
1901 Introductory Hebrew Grammar: Hebrew Syntax. 3rd ed. Edinburgh: T.&T.
Clark.
Dawson, David Allan
1994 Text-Linguistics and Biblical Hebrew. Journal for the Study of the Old Testament Supplement 177. Sheffield: Sheffield Academic Press.
Decaen, Vincent
1995 On the Placement and Interpretation of the Verb in Standard Biblical Hebrew
Prose. Ph.D. dissertation, University of Toronto.
1996 Ewald and Driver on Biblical Hebrew Aspect. Zeitschrift fr Althebristik 9:
12951.
1999 A Unified Analysis of Verbal and Verbless Clauses within GovernmentBinding Theory. Pp. 10931 in The Verbless Clause in Biblical Hebrew: Linguistic Approaches, ed. Cynthia L. Miller. Linguistic Studies in Ancient West
Semitic 1. Winona Lake, IN: Eisenbrauns.
Declerck, Renaat
1986 From Reichenbach (1947) to Comrie (1985) and Beyond. Lingua 70: 30564.
1989 Boundedness and the Structure of Situations. Leuvense Bijdragen 78: 275308.
Dekker, P. J. E., and H. L. W. Hendriks
1994 Situation Semantics. Pp. 395460 in The Encyclopedia of Language and Linguistics, ed. R. E. Asher. Oxford: Pergamon.
DeLancey, Scott
1987 Transitivity in Grammar and Cognition. Pp. 5368 in Coherence and
Grounding in Discourse, ed. Russell S. Tomlin. Amsterdam: Benjamins.
Works Cited
347
Dempster, Stephen G.
1985 Linguistic Features of Hebrew Narrative: A Discourse Analysis of Narrative
from the Classical Period. Ph.D. dissertation, University of Toronto.
Depraetere, Ilse
1995 On the Necessity of Distinguishing between (Un)Boundedness and (a)Telicity.
Linguistics and Philosophy 18: 119.
DeRouchie, Jason S.
2007 A Call to Covenant Love: Text Grammar and Literary Structure in Deuteronomy 511. Gorgias Dissertations in Biblical Studies 30. Piscataway, NJ:
Gorgias.
Dhorme, Paul (Edouard)
191314 La Langue de Canaan. Revue biblique 1011: 36993, 40587.
Diakonoff, I. M.
1965 Semito-Hamitic Languages: An Essay in Classification. Languages of Asia
and Africa. Moscow: Nauka.
1988 Afrasian Languages. Languages of Asia and Africa. Moscow: Nauka.
1990 The Importance of Ebla for History and Linguistics. Pp. 329 in Eblaitica: Essays on the Ebla Archives and Eblaite Language, vol. 2, ed. Cyrus H. Gordon
and Gary A. Rendsburg. Winona Lake, IN: Eisenbraun.
Dik, Simon
1997 The Theory of Functional Grammar, Part 1: The Structure of the Clause.
2nded. Functional Grammar Series. Berlin: de Gruyter.
Dillmann, August
[1899] 1974 Ethiopic Grammar, trans. James A. Chrichton. 2nd ed. Amsterdam:
Philo.
Dobbs-Allsopp, F. W.
20047 (More) on Performatives in Semitic. Zeitschrift fr Althebristik 1720:
3681.
Donner, Herbert, and Wolfgang Rllig
2002 Kanaanische und aramische Inschriften, vol. 1. 5th ed. Wiesbaden:
Harrassowitz.
Downing, Pamela
1995 Word Order in Discourse: By Way of Introduction. Pp. 127 in Word Order
in Discourse, ed. Pamela Downing and Michael Noonan. Typological Studies
in Language 30. Amsterdam: Benjamins.
Dowty, David R.
1977 Semantic Analysis of Verb Aspect and the English Imperfective Progressive.
Linguistics and Philosophy 1: 4577.
1979 Word Meaning and Montague Grammar. Synthese Language Library. Dordrecht: Reidel.
1986 The Effect of Aspectual Classes on the Temporal Structure of Discourse: Semantics or Pragmatics? Linguistics and Philosophy 9: 3761.
Driver, G. R.
1936 Problems of the Hebrew Verbal System. Edinburgh: T.&T. Clark.
348
Works Cited
Driver, S. R.
[1892] 1998 A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions. 3rd ed., with an Introductory Essay by W. Randall Garr.
Grand Rapids, MI: Eerdmans.
Dry, Helen Aristar
1981 Sentence Aspect and the Movement of Narrative Time. Text 1: 23340.
1983 The Movement of Narrative Time. Journal of Literary Semantics 12/2: 1953.
Dryer, Matthew S.
2005 Order of Subject and Verb. Pp. 33437 in The World Atlas of Language Structures, ed. Martin Haspelmath and Hans-Jrg Bibiko. Munich: Max Planck
Digital Library.
Dyk, J. W.
1994 Participles in Context: A Computer-Assisted Study of Old Testament Hebrew.
Amsterdam: Vrije Universiteit Press.
Dyk, Janet W., and Eep Talstra
1999 Paradigmatic and Syntagmatic Features in Identifying Subject and Predicate
in Nominal Clauses. Pp. 13385 in The Verbless Clause in Biblical Hebrew:
Linguistic Approaches, ed. Cynthia L. Miller. Linguistic Studies in Ancient
West Semitic 1. Winona Lake, IN: Eisenbrauns.
Ebeling, Erich
1910 Das Verbum der El-Amarna-Briefe. Beitrge zur Assyriology 10: 3979.
Edelheit, Amos
1944 Blakes Review of Hebrew Grammar: A New Approach by Alexander
Sperber. Journal of Biblical Literature 63: 43839.
Emerton, J. A.
1994a New Evidence for the Use of Waw Consecutive in Aramaic. Vetus Testamentum 44: 25558.
1994b What Light Has Ugaritic Shed on Hebrew? Pp. 5369 in Ugarit and the Bible:
Proceedings of the International Symposium on Ugarit and the Bible, Manchester, September 1992, ed. George J. Brooke, Adrian Curtis, and John F.
Healey. Ugaritisch-Biblische Literatur 11. Mnster: Ugarit-Verlag.
1997 Further Comments on the Use of Tenses in the Aramaic Inscription from Tel
Dan. Vetus Testamentum 47: 42940.
Endo, Yoshinobu
1996 The Verbal System of Classical Hebrew in the Joseph Story: An Approach
from Discourse Analysis. Studia Semitica Neerlandica. Assen: Van Gorcum.
Erbaugh, Mary S.
1987 A Uniform Pause and Error Strategy for Native and Non-Native Speakers.
Pp.10930 in Coherence and Grounding in Discourse, ed. Russell S. Tomlin.
Typological Studies in Language 11. Amsterdam: Benjamins.
Eskhult, Mats
1990 Studies in Verbal Aspect and Narrative Technique in Biblical Hebrew Prose.
Studia Semitica Upsaliensia. Uppsala: Almqvist & Wiksell.
Ewald, Heinrich
1870 Ausfhrliches Lehrbuch der hebrischen Sprache des Alten Bundes. 8thed.
Gttingen: Dieterichschen Buchhandlung.
Works Cited
349
1879 Syntax of the Hebrew Language of the Old Testament, trans. James Kennedy.
Edinburgh: T.&T. Clark.
Exter Blokland, A. F. den
1995 In Search of Text Syntax: Towards a Syntactic Text-Segmentation Model for
Biblical Hebrew. Amsterdam: Vrije Universiteit Press.
Faber, Alice
1997 Genetic Subgroupings of the Semitic Languages. Pp. 315 in The Semitic
Languages, ed. Robert Hetzron. London: Routledge.
Fassberg, Steven E.
1999 The Lengthened Imperative in Biblical Hebrew. Hebrew Studies 40:
713.
Fensham, F. C.
1978 The Use of the Suffix Conjugation and the Prefix Conjugation in a Few Old
Hebrew Poems. Journal of Northwest Semitic Languages 6: 918.
Fenton, Terry L.
1970 The Absence of a Verbal Formation *Yaqattal from Ugaritic and North-West
Semitic. Journal of Semitic Studies 15: 3141.
1973 The Hebrew Tenses in the Light of Ugaritic. Pp. 3139 in Proceedings of
the Fifth World Congress of Jewish Studies, vol. 4: Hebrew and Semitic Language, Folklore, Art and Music. Jerusalem: World Union of Jewish Studies.
Fernald, Theodore B.
2000 Predicates and Temporal Arguments. New York: Oxford University Press.
Fitzgerald, Aloysius
1972 A Note on G-Stem Forms in the Old Testament. Zeitschrift fr die alttestamentliche Wissenschaft 84: 9092.
Fleischman, Suzanne
1990 Tense and Narrativity. Austin: University of Texas Press.
Folmer, Margaretha L.
1991 Some Remarks on the Use of the Finite Verb Form in the Protasis of Conditional Sentences in Aramaic Texts from the Achaemenid Period. Pp. 5678
in Studies in Hebrew and Aramaic Syntax, ed. K. Jongeling, H. L. Murre-Van
Ven Berg, and L. Van Rompay. Leiden: Brill.
1995 The Aramaic Language of the Achaemenid Period: A Study in Linguistic Variation. Orientalia Lovaniensia Analecta. Leuven: Peeters.
Forsyth, J.
1970 A Grammar of Aspect. Cambridge: University of Cambridge Press.
Furuli, Rolf J.
2006 A New Understanding of the Verbal System of Classical Hebrew: An Attempt
to Distinguish between Semantic and Pragmatic Factors. Oslo: Awatu.
Gai, Amikam
1982 The Reduction of Tense (and Other Categories) of the Consequent Verb in
North-West Semitic. Orientalia 51: 25456.
Garr, W. Randall
1998 Drivers Treatise and the Study of Hebrew: Then and Now. Pp. xviiilxxxvi
in A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical
Questions. 3rd ed. The Biblical Resource Series. Grand Rapids, MI: Eerdmans.
350
Works Cited
Works Cited
351
Gordon, Amnon
1982 The Development of the Participle in Biblical, Mishnaic, and Modern Hebrew.
Afroasiatic Linguistics 8/3: 159.
Gordon, Cyrus H.
1938 The Accentual Shift in the Perfect with Waw Consecutive. Journal of Biblical
Literature 57: 31926.
1965 Ugaritic Textbook. Analecta Orientalia. Rome: Pontifical Biblical Institute.
1997 Amorite and Eblaite. Pp. 100113 in The Semitic Languages, ed. Robert
Hetzron. London: Routledge.
Greenberg, Joseph H.
1952 The Afro-Asiatic (Hamito-Semitic) Present. Journal of the American Oriental
Society 72: 19.
Greenstein, Edward L.
1988 On the Prefixed Preterite in Biblical Hebrew. Hebrew Studies 29: 717.
1998 On a New Grammar of Ugaritic. Israel Oriental Society 18: 397420.
2006 Forms and Functions of the Finite Verb in Ugaritic Narrative Verse. Pp.75
102 in Biblical Hebrew in Its Northwest Semitic Setting: Typology and Historical Perspectives, ed. Steven E. Fassberg and Avi Hurvitz. Institute for
Advanced Studies 1. Jerusalem: Magnes / Winona Lake, IN: Eisenbrauns.
Grice, H. P.
1975 Logic and Conversation. Pp. 4158 in Syntax and Semantics, vol. 3: Speech
Acts, ed. P. Cole and J. L. Morgan. New York: Academic Press.
Gropp, Douglas M.
1991 The Function of the Finite Verb in Classical Biblical Hebrew. Hebrew Annual
Review 13: 4562.
Gross, Walter
1976 Verbform und Funktion: Wayyiqtol fr Gegenwart? Ein Beitrag zur Syntax poetischer althebrischer Texte. Mnchener Universittsschriften. St. Ottilien:
EOS.
1981 Syntaktische Ersheinungen am Anfang althebrischer Erzhlungen: Hintergrund und Vordergrund. Pp. 13145 in Congress Volume: Vienna, 1980,
ed.J.A. Emerton. Vetus Testamentum Supplement 32. Leiden: Brill
Hale, Mark
2007 Historical Linguistics: Theory and Method. Blackwell Textbooks in Linguistics. Oxford: Blackwell.
Hamori, A.
1973 A Note on Yaqtulu in East and West Semitic. Archiv Orientln 41: 31924.
Hare, Richard Mervyn
1952 The Language of Morals. Oxford: Clarendon.
1970 Meaning and Speech Acts. The Philosophical Review 79/1: 324.
Harmelink, Bryan L.
2004 Exploring the Syntactic, Semantic, and Pragmatic Uses of
in Biblical
Hebrew. Ph.D. dissertation, Westminster Theological Seminary.
Harris, Zellig S.
1939 Development of the Canaanite Dialects: An Investigation in Linguistic History. New Haven, CT: American Oriental Society.
352
Works Cited
Works Cited
353
Hetzron, Robert
1969 The Evidence for Perfect *Yqtul and Jussive *Yaqtl in Proto-Semitic.
Journal of Semitic Studies 14: 121
Hoffman, G.
1887 Review of Semitic Languages, by Theodor Nldeke. Literarisches Centralblatt fr Deutschland 18: 605.
Hoftijzer, J.
1985 The Function and Use of the Imperfect Forms with Nun Paragogicum in Classical Hebrew. Studia Semitica Neerlandica. Assen: Van Gorcum.
1991 A Preliminary Remark on the Study of the Verbal System in Classical Hebrew.
Pp. 64551 in Semitic Studies in Honor of Wolf Leslau on the Occasion of His
Eighty-Fifth Birthday, ed. Alan S. Kaye. Wiesbaden: Harrassowitz.
Hoftijzer, J., and K. Jongeling
1995 Dictionary of the North-West Semitic Inscriptions. Handbuch der Orientalistik. Leiden: Brill.
Holmstedt, Robert D.
2002 The Relative Clause in Biblical Hebrew: A Linguistic Analysis. Ph.D. dissertation, University of Wisconsin.
2003 Adjusting Our Focus: A Review of Focus Structure in Biblical Hebrew: A
Study of Word Order and Information Structure, by Katsuomi Shimasaki. Hebrew Studies 44: 10113.
2005 Word Order in the Book of Proverbs. Pp. 13554 in Seeking Out the Wisdom
of the Ancients: Essays Offered to Honor Michael V. Fox on the Occasion of
His Sixty-Fifth Birthday, ed. Ronald L. Troxel, Kelvin G. Friebel, and Dennis
R. Magary. Winona Lake, IN: Eisenbrauns.
2009 Word Order and Information Structure in Ruth and Jonah. Journal of Semitic
Studies 59: 11139.
2011 The Typological Classification of the Hebrew of Genesis: Subject-Verb or
Verb-Subject? Journal of Hebrew Scriptures 11. http://www.jhsonline.org/.
Hopper, Paul J.
1979 Aspect and Foregrounding in Discourse. Pp. 21341 in Discourse and Syntax,
ed. Talmy Givn. Syntax and Semantics 12. New York: Academic Press.
1982 Aspect between Discourse and Grammar: An Introductory Essay for the
Volume. in Tense-Aspect: Between Semantics and Pragmatics, ed. Paul J.
Hopper. Typological Studies in Language. Amsterdam: Benjamins.
1987 Emergent Grammar. Pp. 13957 in Grammar and Cognition, ed. Jon Aske, Natasha Beery, and Laura Michaelis. Berkeley Linguistic Society 13. Berkeley:
University of California Press.
Hopper, Paul J., and Sandra A. Thompson
1980 Transitivity in Grammar and Discourse. Language 56: 25199.
Hopper, Paul J., and Elizabeth Closs Traugott
1993 Grammaticalization. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press.
2003 Grammaticalization. 2nd ed. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press.
354
Works Cited
Hornstein, Norbert
1977 Towards a Theory of Tense. Linguistic Inquiry 8: 52157.
1990 As Time Goes By: Tense and Universal Grammar. Cambridge, MA: MIT
Press.
Huehnergard, John
1987a Stative, Predicative, Pseudo-Verb. Journal of Near Eastern Studies 46:
21532.
1987b Ugaritic Vocabulary in Syllabic Transcription. Harvard Semitic Studies 32.
Atlanta: Scholars Press. [Rev. ed., Winona Lake, IN: Eisenbrauns, 2008.]
1988 The Early Hebrew Prefix-Conjugations. Hebrew Studies 29: 1923.
1989 The Akkadian of Ugarit. Harvard Semitic Studies 34. Atlanta: Scholars Press.
1991 Remarks on the Classification of Northwest Semitic Languages. Pp. 28293
in The Balaam Text from Deir Alla Re-evaluated: Proceedings of the International Symposium Held at Leiden, 2124 August 1989, ed. J. Hoftijzer and
G.van der Kooij. Leiden: Brill.
1992 Languages: Introductory Survey. Pp. 15570 in vol. 4 of Anchor Bible Dictionary, ed. David Noel Freedman et al. New York: Doubleday.
2002 Comparative Semitic Linguistics. Pp. 11950 in Semitic Linguistics: The State
of the Art at the Turn of the 21st Century, ed. Shlomo Izre'el. Israel Oriental
Studies 20. Winona Lake, IN: Eisenbrauns.
2004 Afro-Asiatic. Pp. 13859 in The Cambridge Encyclopedia of the Worlds Ancient Languages, ed. Roger D. Woodard. Cambridge: Cambridge University
Press.
2005 A Grammar of Akkadian. 2nd ed. Harvard Semitic Museum Studies 45. Winona Lake, IN: Eisenbrauns.
Huehnergard, John, and Christopher Woods
2004 Akkadian and Eblaite. Pp. 21880 in The Cambridge Encyclopedia of the
Worlds Ancient Languages, ed. Roger D. Woodard. Cambridge: Cambridge
University of Press.
Hughes, James A.
1955 The Hebrew Imperfect with Waw Conjunctive and Perfect with Waw Consecutive and Their Interrelationship. M.A. thesis, Faith Theological Seminary.
1962 Some Problems of the Hebrew Verbal System with Particular Reference to the
Uses of the Tenses. Ph.D. dissertation, University of Glasgow.
1970 Another Look at the Hebrew Tenses. Journal of Near Eastern Studies 29:
1224.
Isaksson, Bo
1987 Studies in the Language of Qoheleth, with Special Emphasis on the Verbal
System. Studia Semitica Upsaliensia. Uppsala: Almqvist & Wiksell.
Izre'el, Shlomo
1978 The Gezer Letters of the El-Amarna Archive. Israel Oriental Studies 8: 1390.
Izre'el, Shlomo, with an appendix by Itamar Singer
1991 Amurru Akkadian: A Linguistic Study. 2 vols. Harvard Semitic Studies 4041.
Atlanta: Scholars Press.
Works Cited
355
Jakobson, Roman
1932 Zur Struktur des Russischen Verbums. Pp. 7484 in Charisteria Gvilelmo
Mathesio qvinqvagenario a discipulis Et Circuli Lingvistici Pragensis soda
libus oblata. Prague: Cercle Linguistique de Prague.
1960 Closing Statement: Linguistics and Poetics. Pp. 35077 in Style and Language, ed. Thomas A. Sebeok. Cambridge: MIT Press.
Janssen, T. M. V.
2006 Montague Semantics. Pp. 24455 in vol. 7 of Encyclopedia of Language and
Linguistics, ed. Keith Brown. Amsterdam: Elsevier.
Janssens, G.
1972 The Present-Imperfect in Semitic. Bibliotheca Orientalis 29: 37.
1980 Review of Hebrisches Perfekt und Imperfekt mit vorangehendem We, by
BoJohnson. Bibliotheca Orientalis 37: 7374.
Jespersen, Otto
1924 The Philosophy of Grammar. New York: Holt.
Johnson, Bo
1979 Hebrisches Perfekt und Imperfekt mit Vorangehendem We. Coniectanea
Biblica: Old Testament Series. Lund: Gleerup.
Johnson, Marion R.
1977 A Semantic Analysis of Kikuyu Tense and Aspect. Ph.D. dissertation, Ohio
State University.
1981 A Unified Temporal Theory of Tense and Aspect. Pp. 14575 in Syntax and
Semantics, vol. 14: Tense and Aspect, ed. Philip J. Tedeschi and Annie Zaenen.
New York: Academic Press.
Joosten, Jan
1989 The Predicative Participle in Biblical Hebrew. Zeitschrift fr Althebristik 2:
12859.
1992 Biblical Weqatal and Syriac Waqtal Expressing Repetition in the Past. Zeitschrift fr Althebristik 5: 114.
1997a The Indicative System of the Biblical Hebrew Verb and Its Literary Exploitation. Pp. 5171 in Narrative Syntax and the Hebrew Bible: Papers of the
Tilburg Conference 1996, ed. Ellen van Wolde. Leiden: Brill.
1997b Workshop: Meaning and Use of the Tenses in 1 Samuel 1. Pp. 7283 in Narrative Syntax and the Hebrew Bible: Papers of the Tilburg Conference 1996,
ed. Ellen van Wolde. Leiden: Brill.
1998 The Functions of the Semitic D Stem: Biblical Hebrew Materials for a
Comparative-Historical Approach. Orientalia 67: 20230.
1999 The Long Form of the Prefixed Conjugation Referring to the Past in Biblical
Hebrew Prose. Hebrew Studies 40: 1526.
2002 Do the Finite Verbal Forms in Biblical Hebrew Express Aspect? Journal of the
Ancient Near Eastern Society of Columbia University 29: 4970.
2005 The Distinction between Classical and Late Biblical Hebrew as Reflected in
Syntax. Hebrew Studies 46: 32739.
2006 The Disappearance of Iterative Weqatal in the Biblical Hebrew Verbal System.
Pp. 13553 in Biblical Hebrew in Its Northwest Semitic Setting: Typology and
356
Works Cited
Historical Perspectives, ed. Steven E. Fassberg and Avi Hurvitz. Institute for
Advanced Studies 1. Jerusalem: Magnes / Winona Lake, IN: Eisenbrauns.
2009 Diachronic Aspects of Narrative Wayhi in Biblical Hebrew. Journal of Northwest Semitic Languages 35/2: 4361.
Joseph, Brian D., and Richard D. Janda
1988 The How and Why of Diachronic Morphologization and Demorphologization.
Pp. 193210 in Theoretical Morphology: Approaches to Modern Lingusitics,
ed. Michael Hammond and Michael Noonan. New York: Academic Press.
Joon, Paul
1923 Grammaire de lHbrew Biblique. Rome: Pontifical Biblical Institute.
2006 A Grammar of Biblical Hebrew, trans. Takamitsu Muraoka. Subsidia Biblica
27. Rome: Pontifical Biblical Institute.
Kahan, J.
1889 ber die verbalnominale Doppelnatur der hebrischen Participien und Infinitive und ihre darauf beruhende verschiedene Konstruktion. Leipzig: Vollrath.
Kamp, Hans
1981 A Theory of Truth and Semantic Representation. Pp. 277322 in Formal
Methods in the Study of Language, ed. Jeroen A. G. Groenendijk, T. M.V.
Janssen, and Martin B. J. Stokhof. Mathmatical Centre Tract 135. Amsterdam:
Mathmatisch Centrum.
Kamp, Hans, Josef van Genabith, and Uwe Reyle
2011 Discourse Representation Theory. Pp. 125394 in vol. 15 of Handbook of
Philosophical Logic, ed. Dov M. Gabby and Franz Guenthner. 2nd ed. Dordrecht: Springer.
Kamp, Hans, and Uwe Reyle
1993 From Discourse to Logic: An Introduction to Modeltheoretic Semantics of
Natural Language, Formal Logic and Discourse Representation Theory.
Boston: Kluwer Academic.
Kamp, Hans, and Christian Rohrer
1983 Tense in Texts. Pp. 25069 in Meaning, Use, and Interpretation of Language,
ed. Rainer Buerle, Christoph Schwarze, and Arnim Von Stechow. Berlin:
deGruyter.
Kanizsa, Gaetano
1979 Organization in Vision: Essays on Gestalt Perception. New York: Praeger.
Kaufmann, Stefan
2005 Conditional Predictions: A Probabilistic Account. Linguistics and Philosophy
28: 181231.
Kautzsch, Emil, ed.
1910 Gesenius Hebrew Grammar, trans. A. E. Cowley. 2nd English ed. Oxford:
Clarendon.
Kawashima, Robert S.
2004 Biblical Narrative and the Death of the Rhapsode. Indiana Studies in Biblical
Literature. Bloomington, IN: Indiana University Press.
Kenny, Anthony
1963 Action, Emotion, and Will. London: Routledge and Kegan Paul.
Works Cited
357
Khan, Geoffrey
2006 Some Aspects of the Copula in North West Semitic. Pp. 15576 in Biblical
Hebrew in Its Northwest Semitic Setting: Typological and Historical Perspectives, ed. Steven E. Fassberg and Avi Hurvitz. Jerusalem: Magnes/ Winona
Lake, IN: Eisenbrauns.
2007 The Morphology of Babylonian Jewish Aramaic. Pp. 10719 in Morphologies
of Africa and Asia, vol. 1, ed. Alan S. Kaye. Winona Lake, IN: Eisenbrauns.
Kiefer, Ferenc
1987 On Defining Modalty. Folia Linguistica 21: 6794.
Kienast, Burkhart
2001 Historische semitische Sprachwissenschaft, mit Beitrgen von Erhart Graefe
(Altaegyptisch) und Gene B. Gragg (Kuschitisch). Wiesbaden: Harrossowitz.
Kittel, Bonnie Pedrotti, Vicki Hoffer, and Rebecca Abts Wright
2004 Biblical Hebrew: A Text and Workbook. 2nd ed. New Haven, CT: Yale University Press.
Klaiman, M. H.
1991 Grammatical Voice. Cambridge Studies in Linguistics. Cambridge: Cambridge University Press.
Klein, G. L.
1990 The Prophetic Perfect. Journal of Northwest Semitic Languages 16: 4560.
Klein, Wolfgang
1994 Time in Language. London: Routledge.
Knudsen, Ebbe E.
1991 Amorite Grammar: A Comparative Statement. Pp. 86685 in Semitic Studies:
in Honor of Wolf Leslau on the Occassion of His Eighty-Fifth Birthday, November 14th, 1991, vol. 1, ed. Alan S. Kaye. Wiesbaden: Harrassowitz.
Knudtzon, J. A.
1892 Zur assyrischen und allgemein semitischen Grammatik. Zeitschrift fr Assyriologie 7: 3363.
1915 Die El-Amarna-Tafeln. Leipzig:
Koehler, Ludwig, Walter Baumgartner, and Johann Jakob Stamm, eds.
19942000 The Hebrew and Aramaic Lexicon of the Old Testament, trans. and
ed.M. E. J. Richardson. 5 vols. Leiden: Brill.
Koffka, K.
1935 Principles of Gestalt Psychology. New York: Harcourt, Brace & World.
Kouwenberg, N. J. C.
2000 Nouns as Verbs: The Verbal Nature of the Akkadian Stative. Orientalia 69:
2171.
2010 The Akkadian Verb and Its Semitic Background. Languages of the Ancient
Near East 2. Winona Lake, IN: Eisenbrauns.
Krahmalkov, Charles R.
1986 The Qatal with Future Tense Reference in Phoenician. Journal of Semitic
Studies 31: 510.
358
Works Cited
Kratzer, Angelika
1995 Stage-Level and Individual-Level Predicates. Pp. 12575 in The Generic
Book, ed. Gregory N. Carslon and Francis Jeffry Pelletier. Chicago: University
of Chicago Press.
Kummerow, David
2007 How Can the Form
Be a Preterite, Jussive, and a Future/Imperfective?
Kleine Untersuchungen zur Sprache des Alten Testaments und seiner Umwelt
8/9: 6395.
2007 On the Biblical Hebrew Verbal System: A Linguistic Critique in Defense of
the Mostly Traditional. Journal of Asia Adventist Seminary 10: 10734.
Krschner, Wilfried
1996 Die Lehre des Grammatikers Dionysios (Dionysios Thrax, Tekhne GrammatikeDeutsch). Pp. 177215 in Ancient Grammar: Content and Context,
ed. Pierre Swiggers and Alfons Wouters. Orbis Supplementa 7. Leuven:
Peeters.
Kurylowicz, Jerzy K.
1961 Lpophonie en Smitique. Prace Jezykoznawcze. Wrocaw: Zakad Narodowy
im. Ossolinskich Wydawn.
1972 Studies in Semitic Grammar and Metrics. Warsaw: Polkiej Akademii Nauk.
1973 Verbal Aspect in Semitic. Orientalia 42: 11420.
Kustr, Pter
1972 Aspekt in Hebrischen. Basel: Reinhardt.
Kuteva, Tania
2001 Auxiliation: An Enquiry into the Nature of Grammaticalization. Oxford: Oxford University Press.
Kutscher, Eduard Yechezkel
1982 A History of the Hebrew Language. Jerusalem: Magnes.
Labov, William
1972 The Transformation of Experience in Narrative Syntax. in Language in the
Inner City. Philadelphia: University of Pennsylvania Press.
Lakoff, R. T.
1971 Ifs, Ands and Buts about Conjunctions. Pp. 11449 in Studies in Linguistic
Semantics, ed. C. J. Fillmore and D. T. Langendoen. New York: Rinehart.
Lambdin, Thomas O.
1971 Introduction to Biblical Hebrew. New York: Scribners.
Lambert, M.
1893 Le vav conversif. Revue des tudes juives 26: 4762.
Leo, C.
1818 An Examination of the Fourteen Verses Selected from Scripture, by Mr.
J.Bellamy, as a Specimen of His Emendation of the Bible. Classical Journal
17: 22140.
Lessau, Donald A., ed.
1994 A Dictionary of Grammaticalization. Bochum: Universittsverlag Dr. N.
Brockmeyer.
Works Cited
359
Levinson, Stephen C.
1983 Pragmatics. Cambridge Textbooks in Linguistics. Cambridge: Cambridge
University Press.
Li, Charles N., and Sandra A. Thompson
1977 A Mechanism for the Development of Copula Morphemes. Pp. 41944 in
Mechanisms of Syntactic Change, ed. Charles N. Li. Austin: University of
Texas Press.
Li, Tarsee
1999 The Expression of Sequence and Non-sequence in Northwest Semitic Narrative Prose. Ph.D. dissertation, Hebrew Union CollegeJewish Institute of
Religion.
Liddell, Henry, and Robert George Scott
1984 Greek-English Lexicon. Oxford: Clarendon.
Lindblom, K.
2006 Cooperative Principle. Pp. 17683 in vol. 3 of Encyclopedia of Language and
Linguistics, ed. Keith Brown. 2nd ed. Amsterdam: Elsevier.
Lindstedt, Jouko
2001 Tense and Aspect. Pp. 76883 in Language Typology and Language Universals: An International Handbook, vol. 1, ed. Martin Haspelmath et al. Berlin:
de Gruyter.
Longacre, Robert E.
1972 Hierarchy and Universality of Discourse Constituents in New Guinea Languages. Washington, DC: Georgetown University Press.
1985 Sentences as Combinations of Clauses. Pp. 23586 in Language Typology and
Syntactic Description, vol. 2: Complex Constructions, ed. Timothy Shopen.
Cambridge: Cambridge University Press.
1990 Storyline Concerns and Word Order Typology. Studies in African Linguistics
Supplement 10. Los Angeles: University of California Press.
1992 Discourse Perspective on the Hebrew Verb: Affirmation and Restatement.
Pp.17789 in Linguistics and Biblical Hebrew, ed. Walter R. Bodine. Winona Lake, IN: Eisenbrauns.
1994 Weqatal Forms in Biblical Hebrew Prose: A Discourse-Modular Approach.
Pp. 5098 in Biblical Hebrew and Discourse Linguistics, ed. Robert D.
Bergen. Dallas: Summer Institute of Linguistics.
1995 Building for the Worship of God: Exodus 25:130:10. Pp. 2149 in Discourse
Analysis of Biblical Literature: What It Is and What It Offers, ed. WalterR.
Bodine. Atlanta: Scholars Press.
1996 The Grammar of Discourse. 2nd ed. Topics in Language and Linguistics. New
York: Plenum.
2003 Joseph: A Story of Divine ProvidenceA Text Theoretical and Textlinguistic
Analysis of Genesis 37 and 3948. 2nd ed. Winona Lake, IN: Eisenbrauns.
Loprieno, Antonio
1986 Das Verbalsystem im gyptischen und im Semitischen: Zur Grundlegung
einer Aspekttheorie. Wiesbaden: Harrossowitz.
360
Works Cited
Lyons, John
1968 Introduction to Theoretical Linguistics. Cambridge: Cambridge University
Press.
1977 Semantics. Cambridge: Cambridge University Press.
Madvig, Johan Nicolai
1895 A Latin Grammar, trans. George Woods. Boston: Ginn.
Marchese, Lynell
1988 Sequential Chaining and Discourse Structure in Godie. Pp. 24773 in Clause
Combining in Grammar and Discourse, ed. John Haiman and Sandra A.
Thompson. Typological Studies in Language 18. Amsterdam: Benjamins.
Matheus, Frank
2011 Ein jegliches hat seine Zeit: Tempus und Aspekt im biblisch-hebrischen Verbalsystem. Kleine Untersuchungen zur Sprache des Alten Testaments und
seiner Umwelt Supplement B/1. Kamen: Hartmut Spenner.
Matthews, Peter
2001 A Short History of Structural Linguistics. Cambridge: Cambridge University
Press.
McCawley, James D.
1993 Everything That Linguists Have Always Wanted to Know about Logic. 2nd ed.
Chicago: University of Chicago Press.
McCoard, Robert W.
1978 The English Perfect: Tense-Choice and Pragmatic Inferences. Amsterdam:
North-Holland.
McFall, Leslie
1982 The Enigma of the Hebrew Verbal System. Historic Texts and Interpreters in
Biblical Scholarship 2. Sheffield: Almond.
Melnik, N.
2010 Modal Predicates in Modern Hebrew. The Proceedings of the 17th International Conference on HPSG. Stanford, CA: CSLI.
Merwe, Christo H. J. van der
1999 The Elusive Biblical Hebrew Term : A Perspective in Terms of Its Syntax,
Semantics, and Pragmatics in 1 Samuel. Hebrew Studies 40: 83114.
Merwe, Christo H. J. van der, Jackie A. Naud, and Jan H. Kroeze
1999 A Biblical Hebrew Reference Grammar. Biblical Languages: Hebrew 3. Sheffield: Sheffield Academic Press.
Mettinger, Tryggve N. D.
1974 The Hebrew Verbal System: A Survey of Recent Research. Annual of the
Swedish Theological Institute 9: 6484.
Meulen, Alice G. B. ter
1995 Representing Time in Natural Language: The Dynamic Interpretation of Tense
and Aspect. Cambridge, MA: MIT Press.
Meyer, Rudolf
1953 Zur Geschichte des hebrischen Verbums. Vetus Testamentum 3: 22535.
1958 Spuren eines westsemitischen Prsens-Futur in den Texten von Chirbet
Qumran. Pp. 11828 in Von Ugarit nach Qumran: Beitrge zur alttestamentlichen Forschung, Otto Eissfeldt zum 1. September 1957 dargebracht
Works Cited
361
362
Works Cited
Moreno Cabrera, J. C.
1998 On the relationship between grammaticalization and lexicalization. Pp.211
27 in Limits of Grammaticalization, ed. Anna Giacalone Ramat and PaulJ.
Hopper. Typological Studies in Language 37. Amsterdam: Benjamins.
Moscati, Sabatino, ed.
1980 An Introduction to the Comparative Grammar of the Semitic Languages:
Phonology and Morphology. Porta Linguarum Orientalium 6. Wiesbaden:
Harrassowitz.
Mourelatos, Alexander P. D.
1981 Events, Processes, and States. Pp. 191212 in Syntax and Semantics, vol.14:
Tense and Aspect, ed. Philip J. Tedeschi and Annie Zaenen. New York: Academic Press.
Muraoka, Takamitsu
1995 Linguistic Notes on the Aramaic Inscription from Tel Dan. Israel Exploration
Journal 45: 1921.
1997a The Alleged Final Function of the Biblical Syntagm <Waw + a Volitive Verb
Form>. Pp. 22941 in Narrative Syntax and the Hebrew Bible: Papers of the
Tilburg Conference 1996, ed. Ellen van Wolde. Leiden: Brill.
1997b Classical Syriac: A Basic Grammar with a Chrestomathy. Porta Linguarum
Orientalium. Wiesbaden: Harrassowitz.
1998 Again on the Tel Dan Inscription and the Northwest Semitic Verb Tenses.
Zeitschrift fr Althebristik 11: 7481.
1999a The Participle in Qumran Hebrew with Special Reference to Its Periphrastic
Use. Pp. 188204 in Sirach, Scrolls, and Sages: Proceedings of a Second
International Symposium on the Hebrew of the Dead Sea Scrolls, Ben Sira,
and the Mishnah, Held at Leiden University, 1517 December 1997, ed. Takamitsu Muraoka and J. F. Elwolde. Leiden: Brill.
1999b The Tripartite Nominal Clause Revisited. Pp. 187214 in The Verbless Clause
in Biblical Hebrew: Linguistic Approaches, ed. Cynthia L. Miller. Linguistic
Studies in Ancient West Semitic 1. Winona Lake, IN: Eisenbrauns.
Mller, Hans-Peter
1983 Zur Geschichte des hebrischen Verbs: Diachronie der Konjugationsthemen.
Biblische Zeitschrift 27: 3457.
1984 Ebla und das althebrische Verbalsystem. Biblica 65: 14567.
1986 Polysemie im semitischen und hebrischen Konjugationssystem. Orientalia
55: 36589.
1991 Wa-, ha- und das Imperfectum consecutivum. Zeitschrift fr Althebristik 4:
14460.
1995a Die aramische Inschrift von Tel Dan. Zeitschrift fr Althebristik 8: 12139.
1995b Ergative Constructions in Early Semitic Languages. Journal of Near Eastern
Studies 54: 26171.
1998 Zu den semitisch-hamitischen Konjugationsystemen. Zeitschrift fr Alt
hebristik 11: 14052.
Newmeyer, Fredrick J.
1998 Language Form and Language Function. Cambridge, MA: MIT Press.
Works Cited
363
Niccacci, Alviero
1987 A Neglected Point of Hebrew Syntax: Yiqtol and Position in the Sentence.
Liber Annuus 37: 719.
1989 An Outline of the Biblical Hebrew Verbal System in Prose. Liber Annuus 39:
726.
1990 The Syntax of the Verb in Classical Hebrew Prose, trans. W. G. E. Watson.
Journal for the Study of the Old Testament Supplement 86. Sheffield: JSOT
Press.
1993 Simple Nominal Clause (Snc) or Verbless Clause in Biblical Hebrew Prose.
Zeitschrift fr Althebristik 6: 21627.
1994a Analysis of Biblical Narrative. Pp. 17598 in Biblical Hebrew and Discourse
Linguistics, ed. Robert D. Bergen. Dallas: Summer Institute of Linguistics.
1994b On the Hebrew Verbal System. Pp. 11737 in Biblical Hebrew and Discourse
Linguistics, ed. Robert D. Bergen. Dallas: Summer Institute of Linguistics.
1995 Essential Hebrew Syntax. Pp. 11125 in Narrative and Comment: Contributions to Discourse Grammar of Biblical Hebrew, ed. Eep Talstra. Amsterdam:
Societas Hebraica Amstelodamensis.
1996 Finite Verb in Second Position of the Sentence: Coherence of the Hebrew
Verbal System. Zeitschrift fr die Alttestamentliche Wissenschaft 108:
43440.
1997 Basic Facts and Theory of the Biblical Hebrew Verb System in Prose.
Pp.167202 in Narrative Syntax and the Hebrew Bible: Papers of the Tilburg
Conference 1996, ed. Ellen van Wolde. Leiden: Brill.
1999 Types and Functions of the Nominal Sentence. Pp. 21548 in The Verbless
Clause in Biblical Hebrew: Linguistic Approaches, ed. Cynthia L. Miller. Linguistic Studies in Ancient West Semitic 1. Winona Lake, IN: Eisenbrauns.
2006 The Biblical Hebrew Verbal System in Poetry. Pp. 24768 in Biblical Hebrew in Its Northwest Semitic Setting: Typology and Historical Perspectives,
ed. Steven E. Fassberg and Avi Hurvitz. Institute for Advanced Studies 1.
Jerusalem: Magnes / Winona Lake, IN: Eisenbrauns.
Nichols, Johanna
2007 What, if Anything, is Typology? Linguistic Typology 11: 23138.
Nldeke, Theodor
[1904] 2001 Compendious Syriac Grammar, trans. James A. Crichton. Reprint. Winona Lake, IN: Eisenbrauns.
Notarius, Tania
2010 The Active Predicative Participle in Archaic and Classical Biblical Poetry: A
Typological and Historical Investigation. Ancient Near Eastern Studies 47:
24169.
Nuyts, Jan
2001 Epistemic Modality, Language, and Conceptualization: A Cognitive-Pragmatic
Perspective. Human Cognitive Processing. Amsterdam: Benjamins.
OConnor, M.
2002 Discourse Linguisitics and the Study of Biblical Hebrew. Pp. 1742 in Congress Volume: Basel, 2001, ed. A. Lemaire. Vetus Testamentum Supplement
92. Leiden: Brill.
364
Works Cited
Works Cited
365
366
Works Cited
1988 Further Remarks on the Hebrew Verbal System. Hebrew Studies 29: 3542.
1990 The Prefix Conjugation Patterns of Early Northwest Semitic. Pp. 40720 in
Lingering over Words: Studies in Ancient Near Eastern Literature in Honor of
William L. Moran, ed. Tzvi Abusch, John Huehnergard, and Piotr Steinkeller.
Harvard Semitic Studies 37. Atlanta: Scholars Press.
199193 Is There Really a Yaqtula Conjugation Pattern in the Canaanite Amarna
Tablets? Journal of Cuneiform Studies 4345: 10718.
1993 The Use of the Precative by Canaanite Scribes in the Amarna Letters.
Pp. 33141 in Mesopotamia Ugaritica Biblica: Festschrift fr Kurt
Bergerhof zur Vollendung seines 70. Lebensjahres am 7. Mai 1992, ed. Manfried Dietrich and Oswald Loretz. Alter Orient und Altes Testament 232.
Neukirchen-Vluyn: Neukirchener Verlag / Kevelaer: Butzon & Bercker.
1996a Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect
Used by the Scribes from Canaan, vol. 1: Orthography, Phonology, Morphosyntactic Analysis of the Pronouns, Nouns, Numerals. Handbuch der Orientalistik. Leiden: Brill.
1996b Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect
Used by the Scribes from Canaan, vol. 2: Morphosyntactic Analysis of the
Verbal System. Handbuch der Orientalistik. Leiden: Brill.
1996c Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect
Used by the Scribes from Canaan, vol. 3: Morphosyntactic Analysis of the
Particles and Adverbs. Handbuch der Orientalistik. Leiden: Brill.
1996d Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect
Used by the Scribes from Canaan, vol. 4: References and Index of Texts Cited.
Handbuch der Orientalistik. Leiden: Brill.
2003a The Suffix Conjugation Pattern in Ancient Hebrew: Tense and Modal Functions. Ancient Near Eastern Studies 40: 342.
2003b The Yaqtul Preterite in Northwest Semitic. Pp. 395407 in Hamlet on a Hill:
Semitic and Greek Studies Presented to Professor T. Muraoka on the Occasion of His Sixty-Fifth Birthday, ed. M. F. J. Baasten and W. T. van Peursen.
Orientalia Lovaniensia Analecta 118. Leuven: Peeters.
Rattray, Susan
1992 The Tense-Mood-Aspect System of Biblical Hebrew, with Special Emphasis on
1and 2 Samuel. Ph.D. dissertation, University of California.
Reichenbach, Hans
1947 Elements of Symbolic Logic. London: Collier-Macmillan.
Reinhart, Tanya
1984 Principles of Gestalt Perception in the Temporal Organization of Narrative
Texts. Linguistics 22: 779809.
Revell, E. J.
1984 Stress and the Waw Consecutive in Biblical Hebrew. Journal of the American Oriental Society 104: 43744.
1985 The Conditioning of Stress Position in Waw Consecutive Perfect Forms in
Biblical Hebrew. Hebrew Annual Review 9: 277300.
1989 The System of the Verb in Standard Biblical Prose. Hebrew Union College
Annual 60: 137.
Works Cited
367
Richardson, M. E. J.
1991 Tense, Aspect and Mood in Ugaritic YQTL. Pp. 28389 in Proceedings of
the Fifth International Hamito-Semitic Congress, ed. Hans G. Mukarovsky.
Verffentlichungen der Institute fr Afrikanistik und gyptologie der Universitt Wien 57. Beitrge zur Afrikanistik 41. Vienna: Afro-Pub.
Ridzewski, Beate
1992 Neuhebrische Grammatik auf Grund der ltesten Handschriften und Inschriften. Heidelberger Orientalistische Studien. Frankfurt: Peter Lang.
Robins, R. H.
1995 The Authenticity of the Tkhnee: The Status Quaestionis. Pp. 1326 in Dionysius Thrax and the Tkhnee Grammatikee, ed. Vivian Law and Ineke Sluiter.
The Henry Sweet Society Studies in the History of Linguistics. 1. Mnster:
Nodus.
1997 A Short History of Linguistics. 4th ed. New York: Longman.
Rogland, Max
2003 Alleged Non-past Uses of Qatal in Classical Hebrew. Studia Semitica Neerlandica. Assen: Van Gorcum.
Romaine, Suzanne
1995 The Grammaticalization of Irrealis in Tok Pisin. Pp. 389427 in Modality in
Grammar and Discourse, ed. Joan Bybee and Suzanne Fleischman. Typological Studies in Language 32. Amsterdam: Benjamins.
Rosn, Haiim B.
1969 The Comparative Assignment of Certain Hebrew Tense Forms. Pp. 21234
in Proceedings of the International Conference on Semitic Studies Held in
Jerusalem 1965. Leiden: Brill.
Rssler, Otto
1950 Verbalbau und Verbalflexion in den Semitohamitischen Sprachen. Zeitschscrift
der deutschen morgenlndischen Gesellschaft 100: 461514.
1961 Eine bisher unbekannte Tempusform im Althebrischen. Zeitschrift der
Deutschen Morgenlndischen Gesellschaft 111: 44551.
1962 Die Prfixkonjugation Qal der Verba Iae nn im Althebrischen und das
Problem der sogenannten Tempora. Zeitschrift fr die Alttestamentliche Wissenschaft 74: 12541.
1981 The Structure and Inflexion of the Verb in the Semito-Hamitic Languages.
Pp.679748 in Bono Homini Donum: Essays in Historical Linguistics. Amsterdam: Benjamins.
Rothstein, Susan
2004 Structuring Events: A Study in the Semantics of Lexical Aspect. Explorations
in Semantics. Oxford: Blackwell.
Rubinstein, A.
1952 A Finite Verb Continued by an Infinitive Absolute in Hebrew. Vetus Testamentum 1: 36267.
Rundgren, Frithiof
1961 Das althebrische Verbum: Abriss der Aspektlehre. Stockholm: Almquist &
Wiksell.
368
Works Cited
Works Cited
369
Sellin, E.
1889 Die Verbal-nominale Doppelnatur der hebrischen Participien und Infinitive
und ihre darauf beruhende verschiedene Construktion. Leipzig: Ackermann
& Glaser.
Seow, C. L.
1995 A Grammar for Biblical Hebrew. 2nd ed. Nashville: Abingdon.
Sheehan, John F. X.
1970 Conversive Waw and Accentual Shift. Biblica 51: 54548.
Shimasaki, Katsuomi
2002 Focus Structure in Biblical Hebrew: A Study of Word Order and Information
Structure. Bethesda, MD: CDL.
Shlonsky, Ur
1997 Clause Structure and Word Order in Hebrew and Arabic: An Essay in Comparative Semitic Syntax. Oxford Studies in Comparative Syntax. Oxford: Oxford University Press.
Shulman, Ahouva
1996 The Use of Modal Verb Forms in Biblical Hebrew Prose. Ph.D. dissertation,
University of Toronto.
1999 The Particle in Biblical Hebrew Prose. Hebrew Studies 40: 5782.
2000 The Function of the Jussive and Indicative Imperfect Forms in Biblical
Hebrew Prose. Zeitschrift fr Althebristik 13: 16880.
Siedl, Suitbert H.
1971 Gedanken zum Tempussystem im Hebrischen und Akkadischen. Wiesbaden:
Harrassowitz.
Silverman, M. H.
1973 Syntactic Notes on the Waw Consecutive. Pp. 16775 in vol. of Orient and
Occident: Essays Presented to Cyrus H. Gordon, ed. H. A. Hoffner Jr. Alter
Orient und Altes Testament 22. Kevelaer: Butzon und Bercker.
Sivan, Daniel
1997 A Grammar of the Ugaritic Language. Handbook of Oriental Studies. Leiden:
Brill.
1998 The Use of QTL and YQTL Forms in the Ugaritic Verbal System. Pp. 89103
in Past Links: Studies in the Languages and Cultures of the Ancient Near East,
Dedicated to Professor Anson F. Rainey, ed. Shlomo Izre'el, Itamar Singer, and
Ran Zadok. Israel Oriental Studies 18. Winona Lake, IN: Eisenbrauns.
Sivan, Daniel, and Zipora Cochavi-Rainey
1992 West-Semitic Vocabulary in Egyptian Script of the 14th to the 10th Centuries
bce. Beer-Sheva: Ben-Gurion University of the Negev Press.
Smith, Carlota S.
1997 The Parameter of Aspect. 2nd ed. Studies in Linguistics and Philosophy. Dordrecht: Kluwer Academic.
1999 Activities: States or Events? Linguistics and Philosophy 22: 479508.
2003 Modes of Discourse: The Local Structure of Texts. Cambridge Studies in Linguistics. Cambridge: Cambridge University Press.
370
Works Cited
2006 The Pragmatics and Semantics of Temporal Meaning. Pp. 92106 in Proceedings, Texas Linguistic Forum 2004, ed. P. Denis, E. Mccready, A. Palmer, and
B. Reese. Somerville, MA: Cascadilla.
2007 Time in Navajo: Direct and Indirect Interpretation. International Journal of
American Linguistics 73/1: 4071.
2008 Time with and without Tense. Pp. 22750 in Time and Modality, ed. Jacqueline Guron and Jacqueline Lecarme. Studies in Natural Language and
Linguistic Theory. Dordrecht: Springer.
Smith, Carlota S., and Mary S. Erbaugh
2005 Temporal Interpretation in Mandarin Chinese. Linguistics 43: 71356.
Smith, Mark S.
1991 The Origins and Development of the Waw-Consecutive: Northwest Semitic
Evidence from Ugarit to Qumran. Harvard Semitic Studies 39. Atlanta:
Scholars Press.
1995 The *qatala Form in Ugaritic Narrative Poetry. Pp. 789803 in Pomegranates
and Golden Bells: Studies in Biblical, Jewish, and Near Eastern Ritual, Law,
and Literature in Honor of Jacob Milgrom, ed. D. P. Wright, D. N. Freedman,
and A. Hurvitz. Winona Lake, IN: Eisenbrauns.
Smyth, Herbert Weir
1956 Greek Grammar. Cambridge: Harvard University Press.
Soden, Wolfram von
1995 Grundriss der Akkadischen Grammatik. 3rd ed. Analecta Orientalia. Rome:
Pontifical Biblical Institute.
Sperber, Alexander
1943 Hebrew Grammar: A New Approach. Journal of Biblical Literature 62:
137262.
1966 A Historical Grammar of Biblical Hebrew: A Presentation of Problems with
Suggestions to Their Solution. Leiden: Brill.
Stassen, Leon
1997 Intransitive Predication. Oxford Studies in Typology and Linguistic Theory.
Oxford: Clarendon.
Steiner, George
1998 After Babel: Aspects of Language and Translation. 3rd ed. New York: Oxford
University Press.
Swiggers, Pierre, and Alfons Wouters
1998 De Tkhnee Grammatike van Dionysius Thrax: De Oudste Spraakkunst in
Het Westen. Orbis Linguarum. Leuven: Peeters.
Senz-Badillos, Angel
1993 A History of the Hebrew Language, trans. John Elwolde. Cambridge: Cambridge University Press.
Talmon, Shemaryahu
1978 The Presentation of Synchroneity and Simultaneity in Biblical Narrative. Pp.
926 in Studies in Hebrew Narrative Art throughout the Ages, ed. J. Heinemann and S. Werses. Scripta Hierosolymitana 27. Jerusalem: Magnes.
Works Cited
371
Talstra, Eep
1978 Text Grammar and Hebrew Bible I: Elements of a Theory. Bibliotheca Orientalis 35: 16875.
1982 Text Grammar and Hebrew Bible II: Syntax and Semantics. Bibliotheca Orientalis 39: 2638.
1992 Text Grammar and Biblical Hebrew: The Viewpoint of Wolfgang Schneider.
Journal of Translation and Textlinguistics 5: 26997.
1997 Tense, Mood, Aspect and Clause Connections in Biblical Hebrew: A Textual
Approach. Journal of Northwest Semitic Languages 23: 81103.
Taylor, John R.
2003 Linguistic Categorization. 3rd ed. Oxford Textbooks in Linguistics. New
York: Oxford University Press.
Testen, David D.
1993 On the Development of the Energic Suffixes. Pp. 293311 in Perspectives
on Arabic Linguistics V: Papers from the Fifth Annual Symposium on Arabic
Linguistics, ed. Mushira Eid and Clive Holes. Current Issues in Linguistic
Theory. Amsterdam: Benjamins.
1994 On the Development of the Arabic Subjunctive. Pp. 15166 in Perspectives
on Arabic Linguistics VI: Papers from the Sixth Annual Symposium on Arabic
Linguistics. Current Issues in Linguistic Theory. Amsterdam: Benjamins.
1998 Parallels in Semitic Linguistics: The Development of Arabic La- and Related
Semitic Particles. Studies in Semitic Languages and Linguistics. Leiden: Brill.
Thacker, T. H.
1954 The Relationship of the Semitic and Egyptian Verbal Systems. Oxford:
Clarendon.
Thompson, Sandra A.
1987 Subordination and Narrative Structure. Pp. 43554 in Coherence and
Grounding in Discourse, ed. Russell S. Tomlin. Typological Studies in Language 11. Amsterdam: Benjamins.
Tigay, Jeffrey H.
1996 Deuteronomy. The JPS Torah Commentary. Philadelphia: The Jewish Publication Society.
Timberlake, Alan
2007 Aspect, Tense, Mood. Pp. 280333 in Language Typology and Syntactic
Description, vol. 3: Grammatical Categories and the Lexicon, ed. Timothy
Shopen. Cambridge: Cambridge University Press.
Tomlin, Russell
1995 Focal Attention, Voice, and Word Order: An Experimental, Cross-Linguistic
Study. Pp. 51754 in Word Order in Discourse, ed. Pamela Downing and Michael Noonan. Typological Studies in Language 30. Amsterdam: Benjamins.
Traugott, Elizabeth C.
1978 On the Expression of Spatio-Temporal Relations in Language. Pp. 369400
in Universals of Human Language, vol. 3: Word Structure, ed. Joseph H.
Greenberg. Stanford, CA: Stanford University Press.
372
Works Cited
Tropper, Josef
1991 Finale Stze und yqtla- Modus im Ugaritischen. Ugarit-Forschungen 23:
34152.
1992 Das ugaritische Verbalsystem. Ugarit-Forschungen 24: 31337.
1995 Das Altkanaanische und Ugaritische Verbalsystem. Pp. 15970 in Ugarit:
Ein Ostmediterranes Kulturzentrum im Alten Orient: Ergebnisse und Per
spektiven der Forschung, ed. Manfried Dietrich and Oswald Loretz. Abhandlungen zur Literatur Alt-Syrien-Palstinas 7. Mnster: Ugarit-Verlag.
1996 Aramisches wyqtl und hebrisches wayyiqtol. Ugaritic-Forschungen 28:
63345.
1998a Althebraisches und semitisches Aspektsystem. Zeitschrift fr Althebristik 11:
15390.
1998b Das Verbalsystem der Amarnabriefe aus Jerusalem. Ugarit-Forschungen 30:
66578.
1999a Die Endungen der semitischen Suffixkonjugation und der Absolutivkasus.
Journal of Semitic Studies 44: 17593.
1999b Tempusmarkierung durch Wortstellung? Zeitschrift fr Althebristik 12:
1046.
1999c Ugaritic Grammar. Pp. 91121 in Handbook of Ugaritic Studies, ed. Wilfred G. E. Watson and N. Wyatt. Handbuch der Orientalistik. Erste Abteilung,
Nahe und der Mittlere Osten. Leiden: Brill.
2000 Ugaritische Grammatik. Alter Orient und Altes Testament. Mnster:
Ugarit-Verlag.
Trubetzkoy, Nicolai S.
1958 Grundzge der Phonologie. Reprint. Travaux du Cercle Linguistique de
Prague. Gttingen: Vandenhoeck & Ruprecht.
1969 Principles of Phonology, trans. C. A. M. Baltaxe. Berkeley: University of
California Press.
Turner, William
1876 The Tenses of the Hebrew Verb. Pp. 338407 in Studies Biblical and Oriental.
Edinburgh: Black.
Van Valin, R. D., Jr., and R. J. LaPolla
1997 Syntax: Structure, Meaning and Function. Cambridge: Cambridge University
Press.
Vendler, Zeno
1957 Verbs and Times. Philosophical Review 66: 14360.
1967 Linguistics in Philosophy. Ithaca, NY: Cornell University Press.
Verkuyl, Henk J.
1993 A Theory of Aspectuality: The Interaction between Temporal and Atemporal
Structure. Cambridge Studies in Linguistics. Cambridge: Cambridge University Press.
Verstraete, Jean-Christophe
2001 Subjective and Objective Modality: Interpersonal and Ideational Functions in
the English Modal Auxiliary System. Journal of Pragmatics 33: 150528.
Works Cited
373
374
Works Cited
Yon, Marguerite
2006 The City of Ugarit at Tell Ras Shamra. Winona Lake, IN: Eisenbrauns.
Zevit, Ziony
1988 Talking Funny in Biblical Henglish and Solving a Problem of the Yaqtl Past
Tense. Hebrew Studies 29: 2533.
1998 The Anterior Construction in Classical Hebrew. Society of Biblical Literature
Monograph Series. Atlanta: Scholars Press.
Zewi, Tamar
1996 The Definition of the Copula and the Role of Third Independent Personal
Pronouns in Nominal Sentences of Semitic Languages. Folia Linguistica Historica 17/12: 4155.
Zuber, Beat
1986 Das Tempussystem des biblischen Hebrisch. Beiheft zur Zeitschrift fr die
altestamentliche Wissenshaft. Berlin: de Gruyter.
Index of Authors
Akman, V. 38
Albright, W. F. 111
Allen, R. L. 3, 4, 5, 9, 10
Allwood, J. 43, 56, 70
Andersen, F. I. 223
Andersen, T. D. 104, 105, 197, 206, 209, 220
Anderson, J. M. 59
Andersson, L.-G. 43, 56, 70
Andrason, A. 149, 174, 178, 182, 184, 200,
220, 221, 245, 246, 251, 263
Anstey, M. P. 202, 218
Anttila, R. 93
Aristotle 1, 5, 7, 19, 20, 57, 59, 180, 181,
194
Arnold, B. T. 313
Austin, J. L. 44, 45, 52, 317
Auwera, J. van der 241
Baayen, R. H. 157, 158, 171, 274
Bach, E. 21
Bache, C. 18, 19, 26, 27, 73, 74, 75, 145,
152, 273
Baker, D. W. 291, 294
Banfield, A. 170
Barco, F. J. del 160, 161
Barnes, O. L. 135, 136
Barr, J. 97
Barwise, J. 38
Bauer, H. 96, 100, 101, 102, 103, 104, 105,
108, 109, 114, 120, 122, 125, 130, 131,
132, 164, 209, 220, 257, 263
Bellamy, J. 85
Bender, L. 95
Bennett, M. 29, 60
Bennett, P. R. 94
Bentinck, J. 161, 163
Bergstrsser, G. 79, 80, 94, 95, 120, 201,
203, 206, 209, 218, 241, 257, 304
Bhat, D. N. S. 121, 200, 205, 260
Bickel, B. 186
Bierwisch, M. 181
Binnick, R. I. 1, 2, 3, 4, 5, 6, 7, 9, 13, 18,
19, 20, 22, 25, 26, 27, 68, 81, 82, 87, 88,
126, 134, 137, 262
375
376
Index of Authors
Index of Authors
Hare, R. M. 44, 45
Harmelink, B. L. 294, 309
Harris, Z. S. 108, 109, 149
Haspelmath, M. 185, 221, 267, 304
Hatav, G. 20, 21, 41, 46, 60, 66, 68, 77, 148,
152, 168, 169, 202, 247, 259, 260, 265,
273, 277, 278, 279, 290, 297
Haug, D. 206
Haupt, P. 99, 100
Heimerdinger, J.-M. 295, 296
Heine, B. 177, 178, 220, 221, 246, 251, 252,
282, 283
Heinrichs, E. 15, 38, 277, 281
Held, M. 108, 215
Heller, R. L. 77, 158, 159, 160, 161, 169,
185
Hendel, R. S. 165, 174, 216, 262, 263
Hetzron, R. 96
Hoffer, V. 83
Hoffman, G. 100, 102
Hoftijzer, J. 223, 240, 241
Holmstedt, R. D. 54, 114, 235, 236, 237, 258
Hopper, P. J. 150, 151, 186, 188, 189, 259,
273, 274, 275, 276, 288
Hornstein, N. 10, 11, 12, 15, 16, 18
Huehnergard, J. 95, 96, 97, 98, 99, 102, 106,
107, 110, 111, 112, 113, 114, 119, 203,
206, 220, 228, 241, 257, 258
Hughes, J. A. 102, 105, 120, 130, 131, 132,
219
Hnnemeyer, F. 177, 178, 246, 251, 252,
283
Isaksson, B. 124
Izre'el, S. 98, 111, 115
Jahn, J. 86
Jakobson, R. 18, 179, 181, 184, 300
Janssen, T. M. V. 37, 55
Janssens, G. 98, 103, 128
Jespersen, O. 4, 5, 6, 7, 8, 9, 13, 14, 27, 42,
179
Johnson, B. 127, 130
Johnson, M. R. 28, 29, 30, 67, 69
Joosten, J. 120, 134, 141, 142, 143, 144,
146, 147, 148, 149, 168, 195, 196, 199,
200, 218, 219, 223, 224, 237, 310
Joon, P. 79, 80, 83, 128, 130, 162, 173,
184, 195, 197, 198, 201, 202, 213, 217,
218, 224, 232, 236, 238, 241, 244, 257,
313, 314
Kahan, J. 223
377
378
Index of Authors
Lyons, J. 18, 43, 44, 45, 49, 52, 53, 75, 122,
140, 170, 179
Madvig, J. N. 3, 4, 5, 6, 7, 9
Matheus, F. 268
Matthews, P. 176, 179, 181
McCawley, J. D. 43, 44, 56, 70, 75
McCoard, R. W. 13
McFall, L. 77, 79, 80, 83, 85, 86, 88, 89, 90,
92, 103, 127, 239, 258, 276
Meillet, A. 186
Melnik, N. 234
Merwe, C. H. J. van der 184, 310, 313
Mettinger, T. N. D. 77, 122, 124, 125
Meulen, A. G. B. ter 38, 39, 40, 277, 278,
281
Meyer, R. 98, 122, 124, 125, 130, 132, 202,
257, 260
Michaelis, L. A. 19
Michel, D. 126, 127, 130, 173, 301
Mill, J. S. 179
Miller, C. L. 199, 267, 291
Mithun, M. 54
Moens, M. 68
Molendijk, A. 151
Montague, R. 37, 38, 55, 56
Moran, W. L. 94, 98, 103, 105, 110, 111,
112, 113, 114, 115, 116, 117, 118, 120,
184, 207, 238, 240, 250, 251
Moravcsik, E. A. 178, 186, 187
Moreno Cabrera, J. C. 188
Moscati, S. 94, 95, 203
Moshavi, A. 235
Mourelatos, A. P. D. 20, 21
Mller, H.-P. 97, 119, 125, 126, 197, 203,
126
Muraoka, T. 119, 130, 162, 173, 201, 226,
229, 244, 313
Naud, J. A. 184, 313
Newmeyer, F. J. 177, 180, 185, 186
Niccacci, A. 133, 152, 156, 157, 158, 160,
171, 244, 272, 158
Nichols, J. 186
Nldeke, T. 120, 201, 250
Notarius, T. 223, 233
Nuyts, J. 50, 51
OConnor, M. 77, 79, 80, 86, 89, 90, 127,
150, 161, 162, 171, 184, 202, 213, 218,
219, 224, 232, 241, 249, 257, 260, 264,
274, 311, 313, 314, 316
Olsen, M. B. 19, 21, 22, 23, 24, 28, 33, 34,
35, 36, 58, 59, 61, 69, 73, 74
Index of Authors
Rllig, W. 251
Romaine, S. 53
Rosn, H. B. 98, 133, 236
Rssler, O. 96, 98, 124
Rothstein, S. 20, 22, 23, 24, 25, 58, 59, 61,
64, 194
Rubinstein, A. 276
Rundgren, F. 122, 123, 124, 125, 130, 140,
143, 144, 123
Ryle, G. 19, 315
Sapir, E. 129
Sasson, V. 119
Satzinger, H. 103
Saussure, F. de 176, 177, 178, 179
Schneider, W. 121, 152, 154, 155, 156, 157,
158, 171, 172
Schoors, A. 124
Schreuder, R. 181
Schroeder, N. W. 84, 85
Schuh, R. G. 199
Schwarzwald, O. 229
Scott, R. G. 57
Searle, J. R. 44, 45, 213
Segal, M. H. 198, 200
Segert, S. 106, 107, 109, 112, 116
Sellin, E. 223
Seow, C. L. 257
Sheehan, J. F. X. 210
Shimasaki, K. 235
Shlonsky, U. 229
Shulman, A. 133, 165, 184, 236, 238, 239,
240, 244, 246, 247
Siedl, S. H. 98
Silverman, M. H. 132, 133
Singer, I. 98, 111, 115
Sivan, D. 106, 107, 109, 111, 112, 116
Smith, C. S. 19, 20, 21, 22, 23, 26, 34, 35,
41, 59, 60, 73, 74, 168, 169, 181, 183,
201, 211, 266, 267, 278, 279, 280, 20,
292, 315, 316, 317, 318, 323, 324, 74, 75
Smith, M. S. 107, 119, 120, 209, 304
Smyth, H. W. 133, 240
Soden, W. von 98, 99, 112, 113, 114, 203,
220, 239, 258
Sperber, A. 82, 219
Stassen, L. 197, 225, 226, 228, 229, 230
Steiner, G. 314
Swiggers, P. 2
Talmon, S. 292, 293, 294, 296
Talstra, E. 150, 152, 154, 155, 157, 171,
172, 230, 272, 304
379
Index of Scripture
Genesis
1:178
1:3243
2:5261
2:6 78, 193
2:10309
2:19294
2:2481
3:1309
3:3246
3:15224
4:1298
4:23133
4:34 159, 297
4:18203
4:26 191, 261
5:24212
6:1191
6:4 218, 226
6:13232
6:17232
7:5312
7:6312
7:7312
8:20263
9:10192
10:8 191, 192
10:24203
11:6191
11:8192
11:30227
14:34298
14:22207
16:8218
16:12309
17:20 203, 214
17:22192
18:7297
18:11 192, 224
18:22231
18:26253
18:33192
19:1142
19:8238
Genesis (cont.)
19:13232
19:19264
19:2829292
21:19319
22:4311
22:16207
22:19242
22:20311
23:1011214
24:1224
24:15192
24:19192
24:22192
24:2324322
24:30231
24:33232
24:39247
24:45 192, 262
25:6 193, 297
25:15262
25:34291
26:10202
26:13192
27:30192
28:15207
29:2248
29:23255
29:3322
29:12294
29:1418290
29:18312
29:21238
29:2325296
29:2830296
29:31227
29:34203
30:7193
30:8203
30:32310
31:9264
31:3278
31:39193
32:5264
Genesis (cont.)
32:6239
33:16296
35:7294
35:15294
3747158
37:2229
37:3 296, 298
37:11298
37:15 79, 218
37:1516268
37:17243
37:30232
39:1298
39:2312
39:1314265
39:22229
40:23159
41:11239
41:49192
41:54 191, 313
42:779
42:37248
43:2192
43:4228
43:10202
43:14207
43:21239
44:33236
45:14296
47:4243
48:10 212, 227
49:33192
50:9227
Exodus
2:10294
3:1229
4:179
4:19294
5:14192
6:2291
7:17231
8:20218
380
Exodus (cont.)
9:13 254, 323
9:1718232
10:4232
10:28241
12 332, 337
12:15333
12:611334
12:1213335
14:8294
15:1 261, 263
15:56260
15:2684
17:9232
17:10296
17:12227
19:12295
19:5305
19:19218
20332
20:1082
20:12249
20:1379
20:1316326
20:1420336
23:5192
24:7314
2530306
2531 319, 332
25:1308
25:13307
25:8308
25:811307
25:1014254
25:11319
29:23193
31:1684
31:18192
33:7193
34:33192
40:36193
Leviticus
15306
Index of Scripture
Leviticus (cont.)
16 319, 332
1:2306
1:3324
1:39321
1:9320
4:2306
4:37307
6:611293
10:1984
16:20192
19:2246
19:9192
23:22192
26:44192
Numbers
1:4749294
7:1192
8:19 239, 240
9:13192
9:18193
16:31192
20:3202
21:17 261, 263
25:1191
27:9252
32:17207
Deuteronomy
1:10212
1:12 245, 248
1:22240
1:30267
2:9241
2:12260
2:31191
3:24191
4:11240
4:22233
4:33265
4:41261
5:23240
5:27203
7:1284
7:22192
8:3289
8:10207
8:19207
9:24229
16:9191
20:9192
23:23192
Deuteronomy (cont.)
26:12192
26:17305
31:3227
31:21261
31:24192
32:8260
32:813118
32:10260
32:11260
32:13260
32:15192
32:29202
32:45192
Joshua
2:8261
2:1516290
2:23289
3:14224
3:15224
3:17224
4:24248
5:5229
7:7202
8:4290
8:24192
8:30 78, 261,
263
9:8 79, 218
11:6232
13:833291
18:8290
19:47264
19:49192
19:51192
22:34226
23:2 212, 224
24:8239
Judges
1:7229
1:25296
1:30192
1:33192
1:35192
2:1260
3:18192
5:1263
6:9240
6:910239
6:10239
6:13226
Judges (cont.)
6:39243
8:1240
8:19202
9:38 78, 218
10:12239
10:18191
11:1294
11:18240
12:3239
13:5191
13:16249
13:23 202, 250
13:25191
14:4231
14:18 202, 250
15:1238
15:2240
15:17192
16:19191
16:21229
16:22191
19:24238
20:31191
20:3147294
20:3940191
1 Samuel
1:1309
1:3194
1:7193
1:10218
1:23195
2:11229
2:15261
2:19193
2:28 239, 240
3:2191
3:3261
3:24267
4:1192
7:13294
7:1516193
7:16193
8326
8:13326
8:23324
8:46328
8:79329
8:1022331
8:11194
9:3261
9:26294
381
1 Samuel (cont.)
10:26305
10:8241
10:13192
11:3253
12:23192
13:10192
13:1718218
14:16294
14:35 191, 263
18:1192
19:11232
20:22207
22:15191
22:23218
23:13192
24:17192
25:34202
28:15239
31:46294
2 Samuel
2:27202
3:6229
4:3294
4:4292
4:7294
4:10239
5:5211
6:18192
7:13310
7:4310
7:9239
7:1117305
7:16228
11:17289
12:8 239, 240
12:2629294
13:12240
13:19159
13:28294
13:2934294
13:34294
13:36192
13:39192
14:5259
15:32193
15:37218
17:11207
17:17193
19:7202
19:30207
20:6248
382
2 Samuel (cont.)
22260
22:12260
22:14260
22:16260
22:24239
23:10218
24:4204
24:14241
24:2223207
1 Kings
1:910159
1:25205
1:41192
3:1192
5:1204
5:25193
6:1129
6:8218
7:13294
7:40192
8:1 261, 263
8:4204
8:54 192, 311
9:1192
11:7 261, 263
11:1522292
12:16204
12:18261
13:1227
14:21204
15:13204
15:16261
15:21192
16:5261
18:15293
18:41314
20:33218
20:39228
22:20305
22:3537294
22:41204
2 Kings
2:9261
2:16248
3:4193
5:6252
6:29294
7:67291
7:1520291
8:27204
Index of Scripture
2 Kings (cont.)
9:33240
10:25192
10:32191
12:11311
13:1320291
13:19202
14:116292
15:5204
15:37191
16:17204
18:34282
18:4 204, 232
18:3681
19:15226
19:23238
20:1281
22:3311
Isaiah
1:9202
1:16192
1:21218
2:2311
2:9265
2:20311
5:13216
5:15265
5:19239
5:25265
6:184
7:1825311
7:20311
7:21312
10:18192
11:9 78, 207,
216
19:7216
24:18265
36:2181
39:1 81, 259
41:484
41:5240
44:1213265
48:18250
49:7264
63:19250
66:7261
Jeremiah
1:1684
4:16265
5:3192
Isaiah (cont.)
7:17230
8:3207
8:6264
8:16265
11:18239
17:18241
18:18241
25:7248
25:1014320
25:29191
26:8192
28:3231
29:14207
29:18207
31:2979
31:36192
32:9239
33:22223
40:15313
43:1192
44:18192
47:1261
48:41216
51:30192
51:63192
Ezekiel
3:3 239, 241
3:1978
9:8239
16:11 239, 240
16:57261
18:279
23:16239
23:20239
31:5192
33:4265
33:6265
43:23192
44:8240
Hosea
2:1218
12:1184
Joel
2:23265
Amos
1:4210
1:4580
1:5210
Amos (cont.)
1:7880
4:78 161, 193
6:3240
7:2192
9:15207
Jonah
1:2242
3:4191
Micah
2:13265
5:2207
Zephaniah
3:2197
Zechariah
3:3229
7:5276
10:6202
11:13239
Psalms
2:2216
3:5299
3:6 239, 299
4:2250
7:5 239, 298,
301
7:7 78, 250
7:13 265, 298,
299
7:16 298, 299,
301
8:6301
9:3243
16:9301
18 260, 298
18:12260
18:14260
18:16260
20:9 265, 299
22:5299
22:30 265, 299
24:2 216, 260
25:2 241, 243
27:13202
28:7299
29:5 298, 301
29:56298
29:6 298, 301
383
Index of Scripture
Psalms (cont.)
29:9 298, 301
29:10298,
301
30:3299
30:12299
31:2241
31:18241
33:9299
34:1298
34:8299
35:21301
36:4192
37:36 298, 299
37:40 265, 298,
301
38:3301
38:13299
38:15301
39:12299
40298
40:2 298, 299
40:23298
40:24298
40:3 298, 299
40:4 298, 299
41:13299
42:6301
43:4238
44:3 298, 299
44:10 298, 299
44:1921298,
299
45:8 299, 301
49:15 265, 301
50:1 298, 301
50:57301
50:6 298, 301
50:16301
50:1618298
50:1718299
52:2298
52:9301
54:2298
55:6301
55:67298
55:7299
55:18 299, 300
59:1298
59:16299
60:1298
60:2298
64:89301
Psalms (cont.)
64:810265,
298
64:10 298, 299
65:9301
65:10299
66:12299
69:11 298, 299
69:1112298
69:12 239, 240,
298, 299
69:15241
69:21 239, 298,
299
69:22298
71:1241
71:21241
73:13 299, 301
73:1314298
73:16 239, 299
75:9299
76:3301
77:7301
77:11299
77:19301
78298
78:10303
78:2425303
78:4451303
78:59303
78:6569302
80:6 298, 301
80:9 299, 300
80:910298
80:10 298, 299
81:8299
81:13299
81:17301
89:20301
89:39301
90:2301
90:3301
90:10 239, 299
92:8301
92:1112301
94:7299
94:22301
94:2223265,
298
94:23299
95:10301
97:4299
97:8 298, 299
Psalms (cont.)
102:5299
102:8299
102:11299
104:31243
104:32299
105298
105:14303
105:1618303
106298
106:2830302
107298
109:3 298, 299
109:4298
109:5299
109:16 298, 301
109:17 298, 299
109:1718298,
301
109:18 298, 299
109:28 265, 298,
299
114:3299
118:14301
118:21299
118:27301
119:26 298, 299
119:52 298, 299
119:55 239, 298,
301
119:59 239, 299
119:5960298
119:67261
119:90 265, 299
119:106239,
298, 301
119:131239,
298, 299
119:147239,
298, 299
119:158239,
298, 299
119:167298,
301
120:1299
136:4299
136:1011300
136:11299
136:18301
138:3299
139:1299
139:5299
139:11301
Psalms (cont.)
143:4299
143:6299
144:3 298, 299,
300
148:14301
Job
1:5 79, 194
1:14229
1:1517239
1:19239
34298
3:6243
3:13202
5:1516265
6:2202
9:34240
10:14253
10:19202
13:21240
14:17265
19:20239
29:17239
31:15240
33:24240
36:7265
Proverbs
7:15264
10:1222
10:12325
11:4223
11:6222
11:8298
11:11222
11:13232
11:31222
12:13298
12:21215
12:25222
13:13222
13:16222
13:20223
14:1215
14:11222
14:17222
15:1222
15:2222
15:2079
16:23222
17:2223
17:20223
384
Proverbs (cont.)
18:1222
18:22298
19:5223
19:9223
20:12215
20:26298
21:11223
21:22 215, 298
22:12298
22:1378
25:4298
26:24222
27:18223
28:18223
29:2222
29:8222
29:23223
Ruth
1:67289
1:9242
1:18192
2:782
2:21207
3:3192
4:9207
Song of Songs
2:7207
Qoheleth
1:3223
1:17 239, 240
1:18223
2:3223
2:16223
2:21223
3:14223
3:15223
3:17223
4:10223
4:11223
4:12223
5:9223
5:11223
5:17223
Index of Scripture
Qoheleth (cont.)
6:7223
6:12223
7:3223
7:7223
7:9223
7:12223
7:18223
7:19223
7:20223
7:26223
8:1223
8:3223
8:5223
8:12223
8:13223
8:15223
9:4223
9:11223
10:1223
10:8223
10:9223
10:12223
10:14223
10:15223
10:18223
10:19223
10:20223
11:3223
11:4223
11:5223
Lamentations
1:13240
1:18227
1:21250
Esther
6:13191
9:23191
Daniel
8:13239
8:15239
8:17239
9:3240
9:34239
Daniel (cont.)
9:4239
9:24192
10:2229
10:16239
10:19239
12:7192
12:8239
Ezra
3:6191
7:28239
8:1517239
8:23239
8:2426239
8:28239
8:31239
9:3239
9:56239
9:6239
Nehemiah
1:4 229, 239
2:1 239, 240
2:6 239, 240
2:9 239, 240
2:13239
4:1191
5:7240
5:78239
5:13239
6:3239
6:8239
6:11240
6:1112239
7:5239
8:13240
12:31239
13:711239
13:9240
13:11240
13:13239
13:17 239, 240
13:19 192, 239,
240
13:21240
13:2122239
Nehemiah (cont.)
13:22240
13:30239
1 Chronicles
1:10 191, 192
1:18203
1:26204
2:4849203
5:2263
13:2240
13:3240
14:15 207, 212
16:2192
17:2228
19:13240
21:2240
21:4204
21:13240
22:5240
27:24191
2 Chronicles
1:10 238, 240
3:12191
4:11192
5:2 261, 263
5:5204
7:1 192, 311
7:12203
10:16204
12:13204
15:16204
16:5192
18:6240
20:9240
20:31204
22:3204
24:11311
26:20204
28:16204
29:29192
31:1204
31:7191
31:10191
34:3191
34:8311