Gansheng Tan1 1Washington University in St. Louis, St. Louis, MO, USA (g.tan@wustl.edu)

Being honest

Writing is not my favorite activity. Yet, writing down my ideas has helped me uncover new ones. In case what I’ve written resonates with others or prompts them to reflectcon often results in more generalized philosophy. Over the years, I’ve explored many forms of writing: fiction, academic articles, diaries, and poems. In this post, I share my personal reflection on a question: What is the purpose of writing?

Writing in childhood and adolescence

In childhood, writing is something we’re taught in school. In adolescence, it becomes a tool for coping with confusion and emotional dissonance. Early adulthood is a period when we are little exposed to the world and societal norms. We encounter requirement and expectation that clash with our biological impulses, for example, being told to show etiquette at the dinner table. During this period of rebellion and reflection, writing seems to be a way to let go some of the grudge, anger, and confusion we have. In this time, we rarely write for approval or understanding. Instead, we tuck these words away, locking them up carefully to protect the more vulnerable parts of our identity.

Academic writing

As we grow older, writing takes on new roles. For artists, it may be a medium for emotional expression. For writers, it becomes a means of reflection, persuasion, or contribution. In academia, where I am in right now, writing is more like a screening tool, than a paid job. In fact, academic writing is rarely compensated directly. There are two main motivation for researchers to write. In socioecologiucal context, all we can expect is that the writing bring back recognition, or adds to a record of contribution, which may lead to job opportunities, or future funding. In addition, there is autrustic human nature to benefit the race as a whole. This impulse continously motivates us to share our research despite the process of writing often takes more time and effort than the research per se.

Outside academia, writing also serves broader social and epistemological functions. It can be a tool to share insight, persuade others, or contribute to a collective knowledge system. The above relates to mechanism of writing, that is, the motivation to write. Following up on this, a second question would be how should we writing. To answer this, I reflect on the function of writing, in the setting of academia.

Writing for reading

Can writing exist without reading? If a piece of writing is never read, not even by its author, what is reason for the existence of such writing ? There are several reasons, for example, emotional coping mechanism, an outlet of emotion, a need for express, or reflection, a tool for integrate thoughts. Writing is motor-cognitive embedment, like hand gesturing during speaking. In the framework of cognitive embodiment, writing might facilitate cognitive reasoning,

Bring judged and viewed is inevitable for academic writing to serve its purpose. For example, the end point of a journal article is to be read. Audience and readership are integrated parts of academic writing. This is dinstinct from writing as a form of art. In the setting where writing is taken as a form of art, writer focus more on what they want to express, for example, criticism, anger, etc. To make the writing understood is less a burden for the writer. And the art often evoked different echoes in reader with different experience. It resonates when a reader has shared a similar experience or can empathize through metaphor and imagery. In contrast, academic writing is designed to inform, instruct, or persuade. As an endeavor in science, academic writing aims to communicate the findings such that people from different walks of life can capture what precise knowledge that the authors want to convey. As a form of communicating knowledge, it prioritizes clarity, logical flow. As a result, the writers now take the task of making reader understand.

Reader-Engineered Writing

Here comes the debates. We can present results and knowedge in a way that only few people or only the writer can understand. That’s a personaly choice. However in today’s life pace, a paper that is difficult to follow, either because of jargon, poor organization, or lack of contextualization, discourages reader from continuing reading and understanding it. This limits its reach and impact. More precisely, such writing can not make little contribution to general knowledge, which violate the autristic purpose of academic reading. A paper that is hard to read for the reader often fail the second objective as well, to get recognition. Peer-review is a unpaid labor. most likely, Your peers will not spend time to understand your work and background before reading your paper, not to mention peer-review do not pay your peer.

Thus, a foundational epistemological principle of academic writing is that it should be readable. This shift introduces a responsibility: to make our message accessible, relevant, and engaging. Finding a right way to communicate knowledge clearly is of course time and effort consuming. Given our limited time, how can we write in an efficient way that convey our knowledge that benefit as many readers as possible? That is why I would like to share a framework for reader-engineereed writing

framework for writing in academic setting

So far, the arguments presented support that considering readership is an important epistemology for academic writing to serve its purpose. A framework built upon natural reasoning process of human cognition can guide academic. Humans want to know why, how, what, and so what. Therefore, a useful structure for the results section—follows the sequence: Motivation → Methods → Results → Interpretation. We begin by asking why the work was done (goal and motivation), then describe how it was done (methods), followed by what was found (results), and finally what it means (interpretation).

Academic writing is not merely a summary of finding. It is an integral part of active research. From my own experience, I will outline how to write to effective communicate knowledge and contribute to research. Research usually starts by writing proposal where goals, background, and methods are pre-defined. This pre-definition serves an epistemic function. It guards against bias and unjustified post hoc claims. Next, the study is conducted according to the protocol, the primary outcome is analyzed using pre-specified methods, sensitivity analyses generate new hypotheses, and writing conveys the gained knowledge.

Once results from planned analysis and necessary controls are available, we should have a sense of the big picture. I advocate drafting the abstract and introduction at this time, before in-depth analysis and sensitivity analyses. This anchors the work in its intellectual context. Results are then organized, with figures used to support and clarify the written narrative. The discussion evolves in parallel with the results: as data are analyzed, insights emerge, patterns are interpreted, and limitations and implications are addressed. Finally, I will describe the methods with two guiding questions:

1. What motivated each analysis?
2. If I were new to this study, what background would I need to replicate the work? Depending on scope, this may be supplemented with detailed tables and appendices. I believe this approach is both efficient and epistemologically sound. Below, I consider three debates that arise within this framework.

Debate 1: Interim Analysis lead to Insight or Bias

One ongoing debate in research design concerns interim analysis. Should we analyze data during the study to optimize resource use and gain early insight, Or should we avoid this to prevent bias and false discoveries?

In conventional paradigms, interim analysis is discouraged because it may lead to “double dipping”, inflating the chance of false positives. However, in domains where data are rare and the cost for experiment is tremendous, such as intracranial recordings in human patients, there is a growing recognition that interim insights may be ethically and practically justified. In this setting, patients volunteer time and accept risk. With limited funding and high logistic cost is limited, it would be a lose if a decade-long study on ten patients yield no conclusion because the paradigm was flawed from the beginning.

Personally, I would highlight the importance of sanity checks at the begining of the study, which hopefully identified potential design flaw. I can see the values of interim analysis in hypothesis generation, but not in hypothesis testing. I would decide abort the study based on the interim analysis, for example, safety concern in a clinical trial. I would design subsequent study based on the interim analysis while following the predefined protocol if no flaw is identified. In addition, carefully documenting the interim results, any changes in hypotheisis, methods allow us gain real-time insight while preserving transparency.

Debate 2: Introduction First, or Last?

One recurring dilemma is whether we should write the introduction before or after analyzing results. Many researchers prefer to start with results and then shape a introduction that frame them more clearly. This approach is practical because most journals favor papers that read compelling, that is, give the impression that the work is advancing the field. Notably, most readers may not be deeply familiar with the field. By shaping the introduction around the results, the paper could stands stronger chance of passing the review process. However, this way may obscure the broader context, risking a missed opportunity to foreground the most important knowledge gap. I recommend the opposite. In my opinion, transparently presenting the big picture in the introduction and refine results to address important knowledge gaps as best as possible does not make a paper less compelling. Research is always constrained by practical and physical realities, so no study can fully resolve every open question. Yet by clearly situating the work within the larger scientific landscape, a paper helps the audience grasp understand the current progress in science, and connects the findings back to the central issues in a frank and authentic way. Such honesty not only respects the reader but also builds trust and emotional resonance with the authors’ motivations, making the contribution feel more meaningful, even if it is necessarily partial.

Writing the abstract and introduction first forces us to re-Review the literature (parallel work emerges during the execution of the project), identify advance during the years when study is conducted, indentify gaps this research can solve, and reorganize results in a way that follows the advance of knowledge in this field,and most importantly, update analyses with most reasonable methodologies and conduct control experiement and statistical control needed to draw conclusions, given newly identified confounds

This approach and mindset not only anchors the analysis but also ensures that conclusions are grounded in existing knowledge. In fact, this approach fit my philosophy of science: to solve unknowns, not by narrating what we’ve happened to find. However I have to admit, a better narration is important to communicate the knowledge to the reader. And the good new is that they do not conflict with each other. A well-articulated story helps readers engage with the work, but it should not come at the cost of obscuring the larger context or the unresolved gaps in the field. Second, although it take times to implement the most appropriate methods and sometimes we can encouter the scenario where the hypothesis defined in the proposal has been somewhat verified, writing the introduction and refining results thereafter might be less efficient than finding a suitable narrative. But in my experience, this often yield higher-impact papers and a smoother peer review experience.

Debate 3: Figures First, or Text First?

Having figures ready early is a common and useful practice when writing journal articles. For many readers, visual content captures attention more readily than text, and figures often provide a coherent storyline on their own. Visually inclined thinkers, in particular, may grasp the main idea more quickly through figures. I do not oppose having the figures themselves represent the full story—ideally, they should stand alone and cohesively convey the main message. That said, I personally prefer to begin with text. As an analytically oriented reader, my understanding from a paper starts with the quantified results and the logic laid out in the text. Figures then serve as a complement representation I turn to when I want to clarify or look for evidence.

Writing the results in text first also helps me determine which aspects of the data are most important to illustrate and what kind of visual representation will best illustrate the knowledge. In this way, figures support the claims and help comprehension through spatial representation. Both approaches are epistemically sound, but my own practice is to write text first.

Conclusion

Writing is time-consuming and rarely rewarded immediately in academia. Yet it is an important component in research, for its ise for reflection, connection, and contribution.

This essay reflects on the purpose of writing, and share a structured and transparent workflow that prompt our writing to fullfil its purpose efficiently: make readers understand.