The Block Method: Writing Scientific Papers Without Tears

Sergey Frolov, University of Pittsburgh

[email protected] ; http://espressospin.org; Twitter @spinespresso

Abstract. This working paper presents a method for writing scientific articles. The idea is to step
away from thinking about a paper as a literary essay with continuous and linear text. Instead,
writing can be done in short blocks which largely do not take into account other blocks. Blocks
can then be assembled into sequences to form a paper. Sequences can be tailored and
re-constituted for different audiences and publication venues. Breaking the daunting task of
writing a paper into short blocks which answer separate questions, and filling out these blocks
with texts, data, figures or calculations, can streamline scientific writing, and save time and
effort. This working paper discusses how to apply the method, how to compose different parts of
the paper (subsequences of blocks), and provides examples as well as an expandable library of
blocktype definitions.

Version 0.34 Date: August 22 2021

Rationale. We write scientific papers the way we do now because not long ago they used to be
printed on actual paper. The PDF format is a digital version of a printed manuscript with its
historic constraints such as splits into pages, multi-panel figures, length restrictions, and literary
requirements such as ‘flowing text’ which is an expectation that each paper is an essay with
smooth transitions between passages, a ‘story arc’ etc. Here we stop and wonder - is such an
illustrated essay really the best form of presenting scientific information?

An author can ‘carry’ an average reader with them only so far into their work due to the
need to get technical. On top of this, scientists rarely read scientific papers from beginning to
end. And if they do, they get stuck a lot because they simultaneously experience the overall
high density of information that obscures the key points and, paradoxically, not enough details to
reconstruct all steps of the discussion. The general trend is to blame the authors for not being
good enough writers. However, while the authors can typically improve their papers by putting
extra work into them, these problems are an inseparable part of the scientific essay-paper
format itself. From the authors’ perspective it looks like an impossible challenge of satisfying
multiple contradicting requirements at once: being clear and eloquent, but presenting all details
of their work accurately, and telling a story with their results which may not be the way they
naturally think about their work.

Another common hassle is that readers, who can be peer reviewers, do not find
answers to their questions even though the authors did answer those somewhere in the paper.
This can, in principle, be improved if the papers were read more carefully, but here again the
problem is fundamental to the linear essay format: it is challenging to keep the entire paper in
your head and to know where to find all the answers.
 With the goal of making it easier to share scientific information in a useful form, the
method described here is an attempt to rethink how we write papers - through stepping away
from the illustrated essay form. By eliminating some of the constraints, the block method cuts
out parts of the process where effort is wasted by everyone involved: authors, referees, editors,
readers - while the accuracy of information and its clarity do not suffer. Paper writing is
reformulated as writing short blocks (few sentences to a paragraph each) and then stringing
them into sequences. By working on each block without worrying about the other blocks the
authors save effort. By choosing which blocks they wish to read the readers find what they are
looking for without reading the whole paper.

The method is based on a realization that many papers contain the same blocks, and in
similar sequences. Moreover, the readers directly look for such blocks and for specific
information in them more often than trying to read the whole paper. For example, a reader may
only wish to look at the figures, or check parameters of a simulation, or the experimental
methods, or they only want to see the summary of results/conclusions, or only want to read an
introduction and abstract. One reason for this is that papers are read with different purposes by
the same reader at different times. Another reason is that papers are read by different
audiences with different levels of knowledge and interest in the work.

The block method can be likened to preparing a PowerPoint presentation where each
slide (each block) addresses a separate point. The logic is clear from the sequence of the
slides, while explicit connections between the slides are not necessary. A future technical
solution can be developed to display a block paper as a deck of slides which can be reshuffled.
At the same time, a block paper can also be printed in PDF and made to look close to the
now-widespread paper format.

One advantage of the method is that the paper can better serve different audiences.
Most of the time you are actually supposed to be writing for several different categories of
readers. For example, a paper should be useful for a beginning student and at the same time for
a senior student, for a professor who works on the same topic and for a professor who is not, for
a journalist, and a funding manager, for a scientist from another field, for your grandma (who
may be a scientist from another field). What often happens is that a traditional essay-paper is a
mad blend of sentences that cater to these different audiences mushed together, and in the end
nobody finds it helpful or easy to read. With the block method you can prepare separate block
sequences for different audiences - even a separate one for your referees. For example, I asked
my undergraduate class to develop 4-block sequences that add to already published papers
and cater to a 4th year student majoring in physics. These four blocktypes can be found in the
blocktype library.

Many authors already write in blocks, sometimes unconsciously. As an exercise, you can
take papers closely related to your work and try to split them into blocks from the blocktype
library. Often it works even if the authors did not intend the paper to be this way!
Basic Principles:

- A Block Method paper is a sequence of short blocks.

- You fill in each block more or less independent of the other blocks.

- Blocks are short: a paragraph or a few sentences per block.

- Figures are also blocks and their descriptions are separate blocks.

- Traditional sections of a paper (Intro, Results, Discussion, Conclusion) are not single
blocks. They are entire sub-sequences of blocks, e.g. The Introductory sequence can
consist of 8 blocks: Impact, Motivation, Context, Previous Work, Challenge, Approach,
List of Results, Future Relevance.

- A library of block definitions (blocktypes) and example block sequences are provided
below. If no suitable blocktype exists you can create a new blocktype and add it to the
library.

- You can add a title to a block. E.g. Context: Quantum Computing. Or you can just label
blocks (e.g. \label command in TeX).

- Blocks can be re-arranged several ways depending on the audience (beginner, expert,
general public), journal (high impact, specialized, preprint), stage of writing (working
notes, draft, final version). This means you can just generate blocks and worry about
their exact placement later.

- No need to worry about linking blocks by phrasing. The text does not need to flow or
connect logically between blocks.Your narrative will be clear from block titles and/or from
their sequence.

- Do not worry about repetition. It is not important if you already said the same thing in
another block. Focus on answering the question posted by the block you are writing (E.g.
‘what is the motivation of this study?’)

- No need to cross-reference blocks, unless it comes naturally. It can take effort and
planning to think about connections between the parts of the paper. You can put all
relevant info in the block itself. (You can edit blocks later if you want to remove repetition,
this is still easier than coordinating between blocks).

- No need for multipanel figures or figure insets. This was used due to constraints of
printed media. Make it one dataset per figure. An exception is when several panels are a
part of a series that must be presented together, or you want to show different
cross-sections of the same data (e.g. linecuts from a 2d array).
- No need to worry about the total length, number of figures, number of references etc.
You can later shuffle some blocks into the supplementary sequence if they do not fit into
your main sequence for any reason, such as journal requirements.

- For a traditional journal, prepare the paper in the Block Method format first (as a
sequence of blocks). Then, some blocks can be merged into paper sections, and their
titles removed. You may want to go through and remove some repetitive sentences at
this stage as well. This is still less work than trying to write a linear essay.

- But you can also try sharing your paper in the Block Method format, i.e. with all blocks
separate, arranged in a certain sequence. You can share it as a LaTeX or a markdown
document. Or perhaps a format that is customized to the Block Method will be developed
in the future.
The Introductory Block Sequence

Introductions of scientific papers are hard to write. It is because Introductions are impossible -
by design. In a few short paragraphs, you are required to review an entire field, cover what was
done before on your question, state the open challenge or otherwise motivate your study,
describe your approach, present your results, and outline the impact and implications of your
work. Sometimes you are compelled to cover 2 or 3 of the above requirements in a single
sentence! All this while simultaneously aiming for experts, beginners, scientists from other fields,
general public, students, professors, editors, referees, funding managers. As a result,
introductions are terrible effort-burners.

The Block Method helps disentangle these different requirements into separate blocks which are
easier to write. The introduction is not a single block but a specialized sequence of blocks,
which can contain as many as a dozen blocks. Because it needs to cover so many disparate
points, it can be thought of as a human face or a wrist: both have a lot of small muscles, each
very important. The Intro sequence, even when short in total text, can easily have the largest
number of blocks and the largest variety of block types.

IMPORTANT. Preparing the Introductory sequence requires that you understand the research
field you are working in. You need to know what other groups have found before you, what is the
reason that you undertook your study, and how your conclusions fit into the knowledge system
of your field. If you are the primary writer of the paper, you need to be on top of these things and
there is no way around it. This understanding comes from reading papers, discussing with your
co-authors and colleagues, attending conferences and seminars. Look at other papers on
similar topics. What ideas do they cover in the introduction? Which papers do they cite? Do you
know all these papers? This is one of the most efficient ways to acquire the relevant background
if you don’t already have it when you start writing.

One time-consuming aspect of writing the introduction is that it typically contains the vast
majority of references, while not dedicating a lot of text to explaining why they are cited. The few
sentences carrying 10-30 citations are very hard to write, because they need to cover a lot of
papers with just a few words, while maintaining scientific accuracy and not offending anyone.
Within the Block Method, you don’t need to cite all papers in the Introductory sequence. You can
have an additional block elsewhere called ‘Further reading’ where you can cite a lot of papers
and explain in detail for each paper (or each group of papers) why they are useful for the reader
to be aware of. This way you are not under pressure to capture all existing literature in your
introduction. You can simply focus on the most relevant references that help the reader get
introduced to your own work. (This point is important: the introduction is written for a truly
interested reader, not for a peer who is only looking for whether or not they were cited.)

An example of a 7-block Introductory sequence is the following. Context--Previous Work--

Challenge--Approach--List of Results---Relation to other works---Impact. Let’s go through these,
first discussing their objectives and then seeing how they work for a ficticios paper on
‘radioactive spin qubits’.

Context - In very broad strokes introduce the general scientific direction you are working in. This
can be very large - ‘climate change’, ‘cancer research’, ‘quantum computing’ etc.

Previous work - A few developments closely relevant to your work.

Challenge - An unresolved question or a widely recognized need for a technological or

conceptual advance. This should be the challenge you are addressing. If there is no challenge
you don’t need to use this block. More on this later.

Approach - How you are addressing the challenge, or more generally, what is your basic method
or idea.

List of results - State your findings, as a simple list. Can be a numbered list or a paragraph of
text. One sentence per finding.

Relation to other works - How your findings differ from, complement or reproduce findings from
other works.

Impact - useful to state explicitly if aiming for a high-impact journal. Can also be replaced with
Future Relevance - consequences of your findings for future works or reinterpretation of past
works.

Now, an example Introductory sequence - for a non-existent paper on quantum computing:

#Context. Quantum computing is a technological paradigm that uses superposition and

entanglement of two-state systems (qubits) to solve certain classes of problems with power that
could scale exponentially with computer size. Quantum hardware required to store and
manipulate qubits comes in many varieties, ranging from photonic circuits to superconducting
nonlinear elements, semiconductor quantum dots and ions suspended in electromagnetic traps.
The current state of the art is tens of qubits connected into mini-processors that can perform
proof-of-principle noisy (large error rate) quantum computation.

#{Previous work}. Among all options, semiconductor qubits have the advantage of scalable
designs analogous to integrated circuits, with the potential to implement thousands or millions of
qubits in the future. The initial experiments were performed on nanostructures made of group
III-V semiconductor materials. Quantum dots occupied by just one electron each were realized
via electrostatic gating in GaAs 2d electron gases [refs]. Single spin readout and manipulation
have been demonstrated and coherence times in the microsecond range have been achieved
with spin control mediated by magnetic as well as electric fields [refs].
#Challenge. However, group III-V semiconductors suffer from the lack of isotopes with zero
nuclear spin. That, and relatively strong hyperfine (electron spin - nuclear spin) interaction, put a
fundamental limit on the maximum attainable dephasing times in group III-V qubits. It is desired
to explore the potential of group IV or group VI semiconductors which can be rid of nuclear
spins completely by isotopic purification, and where natural abundance of nuclear spin is low (as
low as 1%).

#Approach. With this goal, we set out to explore nanostructures based on

Flerovium-Livermorium, a group IV-VI compound previously not studied in the context of spin
qubits. Fl and Lv are very heavy elements which implies strong spin-orbit interaction, an aspect
promising for electric field mediated control of electron spins.

#Results. We use FlLv nanowires to create quantum dots and study their electronic transport
properties. We observe single-electron tunneling onto quantum dots and extract quantum
confinement energies of 200 meV, large enough for room temperature operation. We identify a
limiting factor in that the materials are highly radioactive which makes them hazardous for
humans and also introduces quantum decoherence pathways. We perform numerical modeling
of spin qubit decoherence due to bombardment by alpha particles, which yields good agreement
with measured spin resonance linewidths. We note that since alpha-particles contain an even
number of baryons, their emission does not change the abundance of nuclear spins in the
material.

#Relation to other works. Previously no radioactive compounds were studied as qubit

candidates. However, there has been earlier work suggesting that cosmic rays can interfere with
the operation of a quantum computer [refs].

#Impact. We introduce a new material platform with potential for quantum technologies. A new
decoherence mechanism is identified - due to radioactive decay in the host material. This opens
doors to interdisciplinary work between nuclear science and quantum information science.

Below is an example of an Intro sequence that does not include a Challenge or Impact:

Context - Motivation - Previous Work - Approach - List of Results - Future Relevance

Here, the essential change is the replacement of ‘Challenge’ with ‘Motivation’. You can be
motivated by replicating previous results. You could be looking for an alternative explanation - if
there was no explanation at all this could be formulated as a challenge. You could be loosely
inspired by previous work which gave you an idea. Or you could be not relying on any previous
work and you are introducing a completely new concept (in this case you don’t need a ‘Previous
work’ block or you can use it to state that there is no previous work). Whatever is the case, you
can directly state it in the Motivation block, because people always want to know why you wrote
this paper. Remember that the blocks are largely independent so you can make an approximate
sequence and then adjust it - add/remove blocks or move them around.
Novel blocks to consider including

This section highlights several blocktypes which are not commonly found in papers, but which
can be helpful to both readers and writers, and can facilitate more reproducible scientific
communication. So adding them to papers, whether they use the block method or not, is a good
idea.

Volume and duration of study. A lot can be inferred from the study size. How long did it take?
How many samples were studied? How many datasets were obtained? How many units of
computing power were used? Are the data shown the best, representative, curated or full?
Providing answers to these, and thus defining the size of the study, is central to any paper. Your
work can be impressive simply because of how extensive it is. Also, knowing how common or
unique the observations are is valuable information.

Study design. You can consider bumping your volume and duration of study section up a few
notches and describe how your study was designed. Ideally designs are pre-registered, which
means they are set and announced prior to the start of your work. However, a diagram or text
describing the different stages of the study can be very illuminating. For instance, a study can
have a theoretical and an experimental component. Data selection and processing stages can
be individually described with relevant parameters listed. Below you can find a study design
diagram. The empty squares can be filled with parameters of each step. Such as, how many
datasets were discarded or selected, what statistical technique was applied, etc. Presenting all
these details in a diagram, plus a concise block of Study Design Description, is a low-effort way
of summarizing what you did.
Alternative explanations. When reporting unusual phenomena, it is good practice to be mindful
of other possible explanations. Use this block to comprehensively describe the possibilities that
have occurred to you or those who you discussed your data with. Avoid using one-liner
arguments to dismiss explanations. Go in some detail over each scenario. If truly no alternative
explanations have been identified, use this block to state so. Even if it is one sentence ‘no
alternative explanations have been identified at the time of this writing’. If you have found
several alternative explanations, split them into separate blocks #Alternative1, #Alternative2,
etc.

Further reading. This blocktype can serve as mere citation overflow from the introduction. But it
can also be much more. Think of different readers of your paper, and imagine what they may
want to read as a follow-up or as background. You don’t have any length restrictions, so you can
describe the literature in as many words as you wish, within reason. Provide a few general
review articles and explain which aspects of them are useful for each audience - beginner,
curious outsider, expert that needs to refresh their knowledge etc. List related papers that
contain ideas which overlap with your paper. Or papers that are similarly interesting and useful.
Provide references to methodological papers, textbooks, etc. Another way to use this section is
to include all references that the readers of your preprint and your referees suggested you
include. If you don’t find them directly relevant to other blocks of your paper, you can simply
state that the following references were included upon reader suggestions, or that the readers
thought these would be useful.

Impact. Already discussed in the Introductory sequence part, this block can be used to
formulate why a paper should be published in a high impact journal. Working on this block can
be a good self-assessment exercise. Even if you do not include it in the submitted version of
your paper, you can see how easy/hard it is to explain the high impact aspects of your work and
then show it to a few people to see if they are impressed. In the ideal world this block will
become obsolete, but in the meantime it can save effort at the submission stage. Keep it very
short, 1-3 sentences. If you decide to proceed with a high-impact journal, use this block in your
letter to the editor. Though the referees may also want to know your reasons and they do not
get to see your cover letter.

Senior Undergraduate Sequence

I teach a writing class to Physics Seniors. In this class we have developed a sequence of 4
blocks, which can be added to any paper, even a paper that has already been written. The goal
of this sequence is to create an entry point into your research, and your topic, for a student who
does not have all of the required technical background. The motivation for creating this
sequence in your paper is because senior undergraduate is the career stage at which many
scientists begin engaging in research. The sequence keeps in mind that students are most used
to obtaining new information through classroom instruction which is very different from how the
authors intend their papers. For example, the question ‘what?’ is answered first and then the
‘why?’. Please consider adding this sequence to your papers.

The four blocktypes are (described in detail in the blocktype library):

#{Concept List} - a list of key terms and concepts used, and their short definitions.

#{Explanation of Method Used} - a walkthrough of what the scientists did keeping in mind that a
reader may have never been to any lab.

#{Interpretation of Results} - how the scientists arrived at their conclusion using their method.

#{Broader Context} - what is the significance of the conclusions, and to which field they
contribute.

How to apply the Block Method

1. Start by defining the scope

There are several ways to get going. Most of the time the scope of the paper is known at the
start, but it may not be fully fleshed out. You can start by drafting some of the blocks that deal
with the entire scope of the paper. For example ‘List of Results’, ‘Title’ can be good blocks to
start with. They help you define what the rest of the paper is about, removing a lot of the
uncertainty that makes it hard to start writing. Don’t spend too much time perfecting the Title,
you can come back to it later. Note that ‘Abstract’ may not be the best block to start with,
because in order to write an abstract you need to nail down the phrasing of the conclusions of
the paper, and often the motivation. Plus, different journals have different abstract formatting
requirements so you may need to produce several versions of it.

An efficient way that works for many papers is to start by drafting a set of figures. This includes
diagrams, data, tables, etc. that you want to see in the main text or in the supplementary
sequence. Drafting the figures can help you get on the same page with co-authors about the
scope. It can help you see if your study is complete or if something is missing, and it can help
you formulate the ‘List of Results’. Keep track of how you selected data for the figures and
disclose it in your Study Design or Duration and Volume of Study.

There is no need to finalize the figures at this stage, you can start writing with figures in the
rough form. Given the fact that in the block method you are describing each dataset separately,
you should be able to re-order the figures later with minimal effort. And formatting such as fonts
and colors can take a lot of time, but the text in the figure description is only marginally affected
by it, so no point waiting for perfect figures before you proceed to writing.

Some papers may evolve from technical write-ups in which you were making notes about your
progress. This happens more often with theory or numerical studies. In this case, if you want to
use the block method, consider structuring your preliminary notes in terms of blocks from the
blocktype library. Starting with, e.g. #{Full model description}. If your notes are structured as a
block paper to begin with, you will be able to continue adding blocks as your study progresses
and you acquire final results.

2. Draft the main block sequence with empty blocks (only block names)

It will help you set up the work if you put together a sequence of empty blocks that is to become
your paper. After that, you just need to start filling in blocks, so your work becomes predictable
and rationed. Remember, it does not need to be the final sequence, you can always add or
remove blocks as you go along. You can also define sub-sequences such as Introduction,
Results, Conclusions etc. The sequence that you create depends on the type of study, the
volume of study, and other factors. Examples of sequences are provided below, and you can
contribute your own. Try to keep a modest but steady pace, i.e. fill 1-2 blocks per day and in a
couple weeks your first draft is complete.

3. Fill out the blocks starting with the easiest

For most authors the easiest blocks will be those that deal with the direct description of their
own work. Blocks such as ‘Full/Brief Methods’, ‘Brief/Full Model Description’, or ‘Figure’ and
‘Figure description’ are examples. When filling these blocks, stick to the facts and do not bother
with interpretation, conclusions or analysis. It makes it harder to write when those aspects are
mixed together with the presentation of results.
There is another reason to separate hard results from their interpretation. It can make your
paper appear less trustworthy if you are constantly narrating results with your preferred
interpretation. A reader who does not accept your interpretation will not buy into your data
although they may be correct themselves. A factual description of results is less controversial
and easier to fill in.

4. Fill out the introductory sequence and the interpretation sequence

This is the hardest part to draft. Introduction and interpretation are not identical but they are not
independent. Because in the introduction you need to state your findings which are your
interpretation of your data or calculations.

We talk about the introduction, the hardest part of the paper, above. Here are some
considerations for the interpretation. Interpretation is ideally separate from the presentation of
results. If you find yourself needing to intermix the presentation of results and their interpretation
or discussion, you could do so in separate blocks to form a mixed result/interpretation block
sequence. For instance like this: #Figure 1 - #Figure 1 description - #Figure 1 interpretation.
Description leads the reader through what is in the figure, while the Interpretation discusses the
features in the figure in terms of a scientific hypothesis.

It will be necessary to refer to your results in your Interpretation, so it is helpful to have the
Results sub-sequence already written. Or at least the key features of the data that are relevant
for the Interpretation sub-sequence should be summarized in a dedicated block. You can
re-state your results again in the Interpretation sequence. Repetitions are fine, often helpful to
the reader, and can be managed at a later stage if it becomes necessary to reduce repetition.

Make Alternative explanations a part of your interpretation subsequence, or state that none
were identified.

Another related sequence is the Conclusions. Some people say conclusions are not necessary.
This is because people often just repeat the abstract, and the introduction, for the third time at
the end of the paper. So it feels excessive and redundant. In practice, it is okay to add a
#Conclusions block or a sequence, if for nothing else than just for the feeling of closure that
some co-authors, and referees or readers, may find in this block. What is easiest - add
#Conclusions or fight about this? If you want to make several points in the conclusions part, use
separate blocks for each point - #Conclusion1, #Conclusion2, etc.

One way in which a concluding sequence can enhance the paper and not be a restatement of
the Introduction is if it contains #Future-work or #Future-relevance blocks, or blocks such as
#Limitations - which you can use to disclose the applicability of your findings, the drawbacks of
your approach, and hint at ways to overcome them in the future.

5. Share your drafted blocks and get feedback

Your closest collaborators will typically be happy to help you improve your paper. Within the
block method, you do not need to wait until the entire paper is drafted, you can get feedback on
completed sub-sequences. Since the blocks are very loosely connected, partial feedback can
be meaningful. For instance, you can get feedback on the technical details sequence, such as
‘Methods’ blocks. Or on Figures and their descriptions, or on the Introductory sequence. It may
also be easier for a closely involved co-author to read parts of the paper than the whole paper at
once. But eventually the whole paper needs to be checked and iterated to make sure that all
relevant details are disclosed in one of the many blocks, and that all statements are accurate
and consistent throughout.

6. Fill in the service blocks

This is stuff such as study design, references, acknowledgements, disclosures etc. They are
important but they strongly depend on the rest of the paper therefore it is more efficient to leave
them for later.

7. Write your abstract

Abstract comes first but is easiest to write last! Perhaps last from the main sequence. Once the
paper is essentially finished, you can formulate your abstract based on how the whole paper
turned out. Abstracts are read a lot, sometimes this is the only part of the paper that is read. So
it is worth putting the effort into phrasing it. And when written last, you have the entire paper to
draw from for the abstract.

8. Rearrange and merge blocks as needed

Once you have a complete draft, you may realize that some important points ended up not in
any of the blocks. You can add those discussions as new blocks or add them to existing blocks.
You may also feel that the order of blocks needs adjusting, or some blocks should be moved
between the main and the supplementary sequences.

For many purposes, be it arXiv or journal submissions, you may need to convert your block
method draft into the linear essay form. Maybe in the future if the block method is widely
accepted and people are used to it, you will not need to do this. But at this time most people still
expect to read continuous text not split into blocks, and they may be your referees or editors.

The good news is that when you have a complete draft in the Block Method format, you can
relatively easily merge sub-sequence blocks together, for instance create a single Introduction
section out of a sequence of introductory blocks. And this will bring you pretty close to a linear
essay paper, with less total effort getting there.
It may also be that some arguments get repeated multiple times, because when making blocks
we do not care about repetitions. At this stage you can eliminate repetitions if desired. For
instance, if you ended up expressing the same thought three times, choose the best formulation
or blend the best sentences from different blocks. This is more relevant for linear reading of
papers upon which readers react negatively when they feel that the text is repetetive. Within the
block format repetitions are okay because the paper can be read in different order by different
people, who may also only read parts of the paper.

This all may seem like extra work, but this is still much easier than writing entire sections from a
blank sheet of paper and worrying about how it connects to other parts of the paper. The effort is
not measured in the total length of text written but in how hard it is to come up with text.
Merging, moving and splicing blocks is relatively little effort. Imagine sitting for hours in front of
an empty screen not knowing where to start...
Examples of block sequences

Title - Author List - Abstract

Introductory sub-sequences

Context - Previous Work - Motivation - Approach - Results List - Relation to other Works - Brief
Background

Impact - Context - Previous Work- Challenge - Approach - Results List - Future Relevance -
Brief Methods

Context - Brief Background - Challenge - Approach - Results List - Relation to other Works -
Brief Model Description

Results sub-sequences

Figure 1 (+ Figure name ) - Figure 1 Caption - Figure 1 Description - Figure 2 + Figure Name -
Figure 2 caption - Figure 2 description - Summary of Results

Brief Analytical Result - Brief Derivation - Figure 1 - Figure 1 Caption - Figure 1 Description -
Figure 1 Interpretation - Table 1 - Table 1 Caption - Table 1 Description

Interpretation sub-sequences

Interpretation - Alternative Explanations

Hypothesis Description - Hypothesis Evidence - Alternative Explanation

Conclusions sub-sequences

Limitations - Future Work- Further Reading

Conclusions - Further Reading

Conclusions - Future Work - Future Relevance - Further Reading

Service sub-sequences

Duration and Volume of Study - Data availability - Author Contributions - Funding

Duration and Volume of Study - Data availability - Acknowledgements - Funding - Conflicts of

Interest
Duration and Volume of Study - Data availability - Acknowledgements - Author Contributions -
Funding - Conflicts of Interest

Supplementary sub-sequences

Full Background - Further Reading - Full Derivation - Supplementary Table - Supplementary

Table Caption - Supplementary Table Description

Study Design Diagram - Study Design Description - Full background - Full Model Description -
Supplementary Figure 1 - Supplementary Figure Caption - Supplementary Figure description -
Supplementary Figure Interpretation
The blocktype library

This library contains blocktype descriptions. The idea is to draw blocktypes from the library to
build the block sequence for your paper. The paper is written by filling the chosen blocks with
your material. You do not need to use every block from the library in every paper. You may also
need to use multiple blocks of the same type, e.g. #Figure.

Blocktype: Title
Block description: Title of the paper
What to address: A title that represents the paper, balancing its impact and findings
Comments: Though it is only one sentence, a lot of work and iterations can go into the title.
Come back to it a few times, do not try to write the final title in one go. It is like a very short
poem.

Blocktype: Abstract
Block description: One paragraph explaining the findings, methods and topic.
What to address: This should be primarily a list of your results. However, statements about
motivation and implications have become commonplace, so consider adding them.
Comments: There is often overlap between #Abstract and other blocks in the paper. To save
effort work on it last or once the full scope of the paper’s findings and motivation is fleshed out.

Blocktype: Context
Block description: One paragraph to overview the field of study that this paper belongs to.
What to address: Which general paradigms and very broad ideas define your field. This is
big-picture, the broadest possible paragraph.
Comments: Often the beginning part of the Introduction section. Don’t just use general
sentences, give some examples that illustrate the broad progress and directions. If your work is
at the intersection of different contexts, you can have two context blocks. Mixing contexts into
the same block can make it hard to write and it reduces the readability.

Blocktype: Previous work

Block description: One paragraph on the closely related previous work.
What to address: Focus on explaining what was done before on the same narrow topic. This is
in contrast with #Context which is for a broad overview.
Comments: This is usually placed towards the beginning of the introduction sequence,
especially when linked by logic with #Challenge. But it can just as well be in the back. Do not
talk about the findings of your paper in this block. Use direct sentences referring to particular
works rather than generalized statements. State ‘no closely related previous work’ if this is the
case.

Blocktype: Challenge
Block description: Address in 1-2 sentences a question that has not been previously answered
in your field of study.
What to address: Lack of attention to a particular area, a question asked in previous works but
not answered. Often combines well with #{Previous work}, to define what was already done and
what is still missing.
Comments: Not every work is driven by a challenge, for example a new discovery may not be
an answer to a previous question. Defining a challenge can be a convenient way to impress with
your results without criticizing the work of others - it draws an arrow of progress. This is not a
block where you explain how your work is different from other works, use #{Relation to other
works} for that.

Blocktype: Motivation
Block description: Explain why you undertook this study.
What to address: Try to define whether anything in the previous work (idea, or error?) or
something missing from the previous work motivated your study. Motivation can be not specific,
e.g. this is an interesting area and you decided to apply your method to it.
Comments: #Challenge is one type of motivation, and it covers a lot of cases. It is also fine to
not have a motivation other than pure interest in a problem.

Blocktype: Approach
Block description: A few sentences on your approach to the problem.
What to address: What is your way of addressing the challenge, or of following your motivation.
Do you focus on a particular material, use a certain type of measurement or apply a new
model? Why did you use this approach - what advantage were you hoping to get?
Comments: This is a block that can be especially useful if there are other works in the area that
use different approaches. It can also serve as a bridge between #Motivation and #{Result List}.

Blocktype: Result List

Block description: List your findings as a simple sequence of statements (theses).
What to address: E.g. sentences that could start as ‘we find that…’. Indicate primary and
secondary findings.
Comments: Do not offer a discussion here, don’t draw contrasts, don’t put your work in context,
just state findings. One approach is to reference all main figures of the paper in this block, and
summarize each figure in a sentence.

Blocktype: Relation to other works

Block description: Explain in up to 1 paragraph what you have done the same or different than
specific closely related previous or simultaneous works.
What to address: Similarities or differences in sample preparation, materials used,
measurement techniques, approximations, formalism used etc.
Comments: Use this block only if there are other works on similar topics. This is not about listing
all the things that were not done before or that make your paper new, this is about a fair
comparison. State what the actual similarities and differences are, rather than only saying that
these works are in some way related.

Blocktype: Impact
Block description: 2-3 sentences distilling the breakthrough nature of profound implications.
What to address: Use this block to explain (including to yourself) why this paper should be in a
high impact journal, and be read by many people. Do not include this block if impact is not a
concern.
Comments: It can be a useful exercise to formulate an impact statement even if it is not
required. Note that the impact is not the most important aspect of a paper, but it can be practical
for communicating your results to define the impact nonetheless.

Blocktype: Brief Background

Block description: Provide one paragraph of well-established background necessary to
understand your paper.
What to address: Basic notions for which it is advantageous to present them in your paper, and
not direct readers to a suitable review article or a textbook. Introduce concepts, formalism that
you are going to use extensively, but that has not been developed as part of this study.
Comments: You may want this block in the introductory sequence. You can also add longer
background blocks in the supplementary sequence, and you can make several of them for
different audiences.

Blocktype: Audience Background

Block description: Provide extended audience-specific background.
What to address: At the review article or textbook or popular journal level explain the basic
concepts relevant to understanding your paper in as many details as you are able to put on
paper.
Comments: Define the audience for which you are writing: e.g. faculty, postdoc, senior graduate,
junior graduate, senior undergraduate, scientist from another field, non-scientific audience etc.
You can write several versions of this block for different audiences.

Blocktype: Full Background

Block description: Provide an extensive review-style introduction into your topic
What to address: Key concepts and major previous work milestones that have created your field
and led you to ask the questions that your study is answering, in a long sequence of blocks:
treat each idea here as a separate block.
Comments: This is time-consuming, but - extremely helpful for many groups of readers. This is
similar to an introductory (theory) chapter in a PhD thesis. You may want this in the
Supplementary sub-sequence.

Blocktype: Contents
Block description: List parts of the paper.
What to address: Table of Contents. You can list all the blocks, or you can list sub-sequences,
e.g. Introductory, Model Description, Results, etc.)
Comments: Useful for longer papers. Often generated automatically.

Blocktype: Brief Experimental Methods

Block description: One paragraph explaining techniques, but not going into the full technical
details.
What to address: Write it as a list of methods used or of procedural steps. Keep it short.
Comments: Use Full Methods for complete description.

Blocktype: Full Experimental Methods

Block description: Provide the technical details of your work, in the greatest detail possible.
What to address: Give all parameters of the processes and techniques you used. Think about
what question a reader may have and answer those.
Comments: This should be a long block.

Blocktype: Brief model description

Block description: Describe in 1-2 paragraphs the essence of the model you are using or that
you have created.
What to address: 1-2 equations that are key for this paper, e.g. your model Hamiltonian for a
physics theory paper, definitions, approximations used and the way you solve the model.
Comments: Focus on the conceptual discussion and reasons why you chose this model.

Blocktype: Full model description

Block description: Describe in sufficient detail the model you are using.
What to address: Model at the level that would allow the reader to reproduce your calculations.
Always include your code with the paper ideally on a server where the code can be executed -
or provide instructions for how to run your code.
Comments: This should be a long block, or a block sequence.

Blocktype: Concept List (Undergraduate)

Block description: A glossary of terms that are found in the paper, and their brief definitions.
What to address: Help a reader without the full background connect to what they know and get
them through the large volume of technical vocabulary in your paper.
Comments: You can write this for other audiences too.

Blocktype: Explanation of Method Used (Undergraduate)

Block description: Explain what the researchers did and how they did it.
What to address: Focus on the process used and the raw findings, e.g. what instruments
showed which measurements. Imagine an audience that does not know what your lab or your
process looks like.
Comments: You can write this for other audiences too, and you can use the concepts from the
#{Concept list}.

Blocktype: Interpretation of Results (Undergraduate)

Block description: Explain which scientific hypothesis the researchers explored and how it
connects to what they did, as described in the #{Explanation of Method Used} block.
What to address: Explain to a senior undergraduate how the findings of the study (#{List of
Results}) connect to the method used. This part is often skipped and is not obvious unless you
are already a practicing scientist.
Comments: Here you can use concepts from the #{Concept list}.

Blocktype: Broader Context (Undergraduate)

Block description: Explain how the results and their interpretation fit into a broader research
area.
What to address: Do results have immediate implications, do they indicate progress towards a
goal, do they open new research directions, do they have practical applications etc.
Comments: This is similar to the #Context blocktype but it is read in reverse order. #Context is
often one of the first blocks, while this undergraduate block connects the ‘what’ to the ‘why’. The
idea is to build a bridge from the classroom style where material is provided first and then put in
context to enhance understanding.

Blocktype: Figure (or Table)

Block description: A visual display itself.
What to address: Present data accurately and clearly.
Comments: There are many rules and guides for how to make figures. One idea that The Block
Method advocates is to avoid multipanel figures, or insets. Unless panels are useful to present a
sequence that must be displayed together, make separate figures. Combine if the journal where
you want to submit limits the number of figures.

Blocktype: Figure (or Table) Caption

Block description: A very brief statement of what is in the figure.
What to address: Axis labels, variables plotted, relevant parameters and conditions.
Comments: Avoid including interpretation or discussion of data in the caption, save that for
#{Figure Description} or #{Figure Interpretation}.

Blocktype: Figure (or Table) Description

Block description: Describe the figure or table as if you are talking to a person.
What to address: Walk the reader through your visual display, what do we see? Focus on the
data itself and your chosen way to present it. Explain how you chose the data for this figure and
what other data show - are there more examples in other figures or in the supplementary?
Comments: Separate interpretation of your data (e.g. comparison to your chosen hypothesis)
into a different block or into a dedicated Interpretation sub-sequence that follows the Results
sub-sequence.

Blocktype: Figure (or Table) Interpretation

Block description: Compare the data in a given Figure (Table) to a model or hypothesis.
What to address: Your analysis of the data, and anything you infer from the data. Discuss
limitations of your interpretation, here or in a separate block.
Comments: If you find it more convenient to discuss several Figures together, first provide an
interpretation-free Results sub-sequence based on those figures, and then insert an
interpretation block.

Blocktype: Analytical results (e.g. a formula, a theorem)

Block description: For a derived analytical result, present the outcome as a separate block.
What to address: Include a formula, with description, but skip derivation.
Comments: It will be helpful for many readers who want to interact with your result (use your
formula), but do not need to follow its derivation. Can be accompanied by its own #{Analytical
Result Description} or #{Interpretation}.

Blocktype: Full Derivation of Analytical Result

Block description: Provide a detailed derivation of the analytical result
What to address: Your full derivation or proof, in as many details as possible. State all
assumptions made (can also be a separate block).
Comments: Can also include #{Brief Derivation} with only the key steps or ideas listed. Label
each derivation when multiple derivations are included.

Blocktype: Result (Not contained in a figure)

Block description: Use for a result that you only need text to explain.
What to address: A factual statement of a finding made during the study. Can be a
generalization over many samples. A negative finding, e.g. ‘we find no effect’.
Comments: This can be accompanied by its own #Interpretation block.

Blocktype: Summary of Data

Block description: A factual summary of results presented in Figures, Tables etc.
What to address: Can contain any number of very technical details, as long as they are
described in the results sequence. This is the difference with the #{Results List}. Another
difference is that this block does not include any interpretations.
Comments: Can be useful at the end of the Results sub-sequence. Can be referred to from the
Interpretation sub-sequence.

Blocktype: Hypothesis Description

Block description: State in one one paragraph what is the hypothesis you are investigating.
What to address: Purely the hypothesis, without wading into the evidence towards or against it.
When comparing multiple hypotheses this can be developed into a sequence, e.g. Hypothesis1,
Hypothesis2, Hypothesis1_vs_2.
Comments: it is good to have this be separate from your results, so that you avoid narrating
your results with your preferred hypothesis. When the hypothesis is separated into a dedicated
block you are well set up to compare your results and point out where the hypothesis fits and
where it does not, in additional blocks.

Blocktype: Hypothesis Evidence

Block description: Draw evidence from the results sequence, e.g. from #{Summary of data} and
explain how it relates to the hypothesis you described.
What to address: You can present supportive evidence, as well as negative evidence, using
separate blocks. Does the hypothesis fit perfectly? Really? Are you sure?
Comments: It is cleaner, and also easier, to separate out the hypothesis description and
hypothesis evidence into separate blocks, and to present results in their own dedicated blocks.

Blocktype: Interpretation
Block description: Provide interpretation of your results. Use where #Hypothesis blocks do not
work. Also known as ‘Discussion’.
What to address: Anything inferred from the results, as opposed to results themselves. This can
be a long block or a sequence of blocks of the same blocktype.
Comments: Interpretation can also be many separate blocks intermixed with figures, see
#{Figure Interpretation}. It is easier and cleaner to have this separate from the presentation of
results themselves. If the interpretation of your experiment comes in part from a theory model,
present the model and its results separately, e.g. using #{Model Description} and appropriate
figures and results blocks. One type of #Interpretation is #Comparison, e.g. between experiment
and theory.

Blocktype: Comparison
Block description: Use to compare model to data, compare different hypotheses, etc.
What to address: What are the similar features of two bodies of data, data vs model, of different
interpretations and hypotheses. Where do the two, or more, ideas diverge or differ?
Comments: This is a common form of providing interpretation of results, e.g. to compare data
and model.

Blocktype: Alternative explanations

Block description: Explain all ideas that may contradict your interpretation.
What to address: Is there another explanation? How does it compare to your chosen
interpretation? Even if no well-defined alternative exists, does your interpretation fully capture
your results?
Comments: This is a more universal block than would be ‘negative hypothesis evidence’. It is
recommended to add this to every paper, even if there truly are no alternative explanations -
state ‘no alternative explanations have been identified at the time of this writing’.

Blocktype: Limitations
Block description: Explain any limitations of your overall approach or of your given study.
What to address: Why do the results not match the hypothesis? Does the chosen method have
limitations that dictate accuracy? Are approximations made likely to discard some effects? Were
samples/specimens/datasets used limiting the accuracy/comprehensiveness or
theory-vs-experiment match? Are the limitations fundamental or are they study/sample specific,
i.e. specific to your attempt?
Comments: Perhaps painful to write, from the ego point of view, but amazing to do and kind of
liberating once included. Gives you another chance to present your results in an fair and
objective way. If you connect this block to #{Future Work}, you can explain how some of the
limitations can be overcome in the future.

Blocktype: Future Work

Block description: Explain what subsequent studies are possible because of, or in addition to
your study.
What to address: Next major steps along a path of research; follow-on works focused on
unanswered questions; technology or methods developments that may take place to resolve
limitations.
Comments: Create a list of items.

Blocktype: Future relevance

Block description: Explain how your work can be relevant in the future.
What to address: Starting with your list of results, show how some of them may be used in other
works, how they can enable new research direction or technology, new ways of thinking about
problems or new ways of solving problems.
Comments: Overlaps with #Impact and #{Practical applications}. But is more precisely defined.
Goes well with #{Future work} and can substitute #Conclusions. Different from #{Future Work}
because it focuses on your results not on future results.
Blocktype: Conclusions
Block description: State the conclusions of your work.
What to address: Your results.
Comments: Usually repeats the #{Results List} from the Introductory sub-sequence, sometimes
rephrased. Add it if it makes you feel better...

Blocktype: Practical applications (Broad Audience)

Block description: Explain in the broadest possible terms how your work can be useful in
everyday life. Or state that it is not.
What to address: Can it lead to new technologies, medicines and cures, deeper understanding
of the world, safer world, more just world? Explain these using no jargon, for a very broad
audience, e.g. CNN watchers or New York Times readers.
Comments: Important to not get carried away. Society will check back with you about the flying
cars you promised, in a while. This overlaps with #{Future Relevance} but is different because it
is not aimed at experts.

Blocktype: Further reading

Block description: Reference other work (or groups of papers) and explain their relevance to
your study.
What to address: Are there helpful review articles? What should a novice reader look up? Can
you point to more technical articles (for experts)? Articles on similar ideas? References
suggested by readers, for whatever reason? Use separate sentences to describe separate
papers, treat it as a list.
Comments: Many readers would appreciate this. And it is better handled in a separate block
than in a crowded introduction.

Blocktype: Duration and Volume of Study

Block description: One paragraph to define how extensive the reported study was.
What to address: Did it take 10 years or a week? Have you collected 10,000 datasets or 5? How
many samples have you prepared, subjects interviewed, computing units used?
Comments: There are no wrong answers, yet to experts this information will be helpful in
understanding, evaluating your work. And in comparing it to other studies or informing their own
work.

Blocktype: Study Design Diagram

Block description: A workflow diagram that goes over the main steps of your study, and contains
details of each step.
What to address: Which steps were involved in the preparation of specimens? Which steps
were involved in acquiring data? How were the data curated, processed and selected for
figures?
Comments: Rectangles connected by squares, see an example in this working paper, or here:
https://www.nature.com/articles/s41567-018-0342-2 . Can include all information from the
#{Duration and Volume of Study}, and provide more details broken down step by step.

Blocktype: Study Design Diagram Description

Block description: Explanation of each of the elements of the #{Study Design Diagram}.
What to address: Give parameters and detailed sub-steps for each major step. Can draw upon
the Methods blocks, and #{Duration and Volume of Study}.
Comments: Study Design can be defined post-factum for every study, and the diagram with
detailed description can be an efficient and low-effort way of laying out what you did in a
concise, visual and informative way. Study Design can also be drafted prior to the
commencement of the study, if practical for your field and your type of study.

Blocktype: Availability of data and code

Block description: Explain how original source and/or processed data and code from your study
can be obtained by readers.
What to address: Where are the data and code archived? What amount of data is available? -
only data presented in Figures, full data, curated data?
Comments: If data are unavailable, state so. Avoid saying that data are available upon request,
especially ‘reasonable’ request. If you don’t mind sharing data at some point, share it up front.
Repositories such as Zenodo now offer 50Gb of space per record and a DOI for each record.

Blocktype: Author List

Block description: List all the co-authors and their affiliations.
What to address: Who is the first author? Who is the last author? Who is an in-between author?
Who contributed equally? Who is the corresponding author?
Comments: R.U. Kidding from the University of Penguinsburg has contributed to your work.

Blocktype: Author contributions

Block description: Explain what each co-author did during the study.
What to address: Who wrote the paper? Who performed simulations? Who prepared samples?
Who performed measurements? Who supervised? Who assisted in another significant way?
Comments: First introduced by high-impact journals hit by fraud scandals, this block is actually
rather meaningful especially when several groups have collaborated on one paper. Indicate
corresponding authors, responsible for future correspondence about this paper.

Blocktype: Acknowledgements
Block description: List persons or facilities that helped you but not enough to be named
co-authors.
What to address: Try to specify what they helped with rather than generic ‘technical support’ or
‘useful discussions’, if appropriate to disclose.
Comments: Sometimes grants are lumped in here, but really that should be a separate block.
Blocktype: Funding Sources
Block description: List grants used to support this work.
What to address: Names of funding agencies, grant numbers.
Comments: This is very important, obvi. Ask your co-authors for their funding information.

Blocktype: Conflicts of Interest

Block description: Disclose financial or other conflicts.
What to address: Was the study funded by a company that benefits from outcomes? Is there a
patent pending?
Comments: Rarely comes up in many disciplines, but do ask yourself if you have anything to
disclose here. For example, if you have circumstances that preclude free sharing of data, they
can be explained here. Unless stated, journals and agencies will consider there to be none.