A critical role for author voice in scientific writing

Written by: Peter van Gelder
Published: 22 December 2024

The final talk in the 2024 UniSIG programme featured Dr Kate Mc Intyre’s presentation of ‘A critical role for author voice in scientific writing’. Kate is the in-house scientific editor at the Department of Genetics, University Medical Center Groningen and has had extensive experience as a researcher and writer of scientific texts. The presentation took place on Zoom and was attended by more than 20 participants. Kate articulated her position on the role of voice in scientific writing and the need to nurture its development, especially in junior research writers. The presentation lasted for about 40 minutes and was followed by a short period of questions and comments from participants.

Why does ‘voice’ matter?

Kate started by explaining her motivation for focusing on author voice in scientific writing, and offered two key factors. The first was her ‘gut feeling’ that, in the face of AI generative tools, the importance of personal voice is a key issue; the second was her overall impression that voice does matter based on her own experience of editing, teaching, and writing scientific texts. However, acquiring and purposefully using that voice can prove difficult for many scientists as they are neither natural writers nor have English as their native language. Moreover, to take a stance and use voice to project a position can lead to criticism from peers, which can be especially challenging for junior members of a team. In a recent poll of researchers in Kate’s department, many named ‘writing’ as what they least like about their job. Kate believes that this attitude, especially prevalent among junior researchers, can be linked to their struggles to find a personal voice that is independent of the more experienced voices in their research teams.

What is voice?

Kate quoted several definitions of voice, which she condensed into two essential elements: the reflection of self that is expressed in text and the consequent creation of identity, agency and power. Kate then explored two theoretical frameworks of voice in academic texts by introducing Ken Hyland’s stance and engagement model (2005) and Sachiko Yasuda’s more recent model of concepts and categories signifying author voice (2022). Hyland’s model accentuates interaction as he divides the framework into two sections: the first contains the linguistic elements by which writers communicate stance through voice and the second contains the ways that writers engage directly with their readers. Yasuda’s framework addresses the various concepts that convey voice in three categories: voice at the linguistic level, voice at the whole text level, and voice at the non-linguistic level.

How is voice expressed in scientific writing?

We next looked at how voice is expressed in scientific writing, exploring three areas of choice by which authors can use an array of tools to produce personal voice to engage with their readership. The first area of choice is what authors choose to say (and not say). This includes choosing what to describe in the research gap and its relevance to the field/world problem to provide motivation, choosing what previous research to include (or not), choosing what experiments/results to include in the main text as opposed to those moved to supplementary material, and choosing what (and to what extent) to discuss in the Discussion section. The second area of choice is how authors can use linguistic devices to express their voice. This includes word choices (hedges, boosters, attitude markers), choices of how ideas are presented and structured, and the choice of what to say when introducing other research (whether to support a general observation or to explicitly name or express an opinion of a study). The third area of choice deals with what and how to choose the schematics, data visualizations and tables, and how to treat the data and statistics that describe them.

Author’s voice vs field-voice

Kate then introduced the bigger picture of how author voice is expressed and modified within the larger ecosystem of other voices present in the process. In the collaborative journey from writing a scientific text to having it published, the predominant voice may start with the primary author(s) but, along the way, there are others who can exercise a ‘field-voice’. These include co-writers (both supporting and passive), supervisors and senior authors, journal editors and reviewers as well as peers in the research field, all of whom can exercise their voice to influence and merge the original author voice into a field-acceptable voice. Although the inexperienced primary author may be aware of the need to merge their own story with the field-specific voice, they struggle to maintain their personal voice as the writing develops. In the face of this erosion and merging of their personal voice and the pressure of being able to navigate scientific storytelling, how does this affect junior researchers? Kate gives their answer: ‘It can be hard to write when you can feel everyone breathing down your neck’.

Problems and solutions

How can we, as teachers and editors of scientific writing, help science researchers and writers, especially the junior members, develop their personal voice? How can we help them create a narrative in which they are able to express their ideas and, at the same time, merge their voice with that of their research community? The answers to these questions begin with first understanding the problems. Kate presented us with five problems and their solutions.

Problem 1. Very linear storytelling

   Problem – Author can only tell the story in their way
      • Struggles to merge IMRaD structure with their narrative
      • Trouble with transitions and paraphrasing
      • Tendency to repeat factual statements in total, often multiple times
      • Trouble with self-editing

   Solutions
      • Lessons on
        − transitions and paraphrasing
        − how to summarize and reference
        − how to refer to what was said earlier in a text
      • Explanation of reader needs

Problem 2. Navigating a ‘big swerve’

   Problem – Research did not go as planned/hoped
      • Disappointment/frustration creeps into the text
      • Lackluster prose and language, devaluing their own work
      • Failure to adequately address important (if disappointing) findings

   Solutions
      • Show them how to edit out emotion markers
      • Propose ways to emphasize the importance of findings

Problem 3. Navigating a ‘side-quest’

   Problem
      • Author struggles to tell a story that deviates from the overall narrative

   Solutions
      • Be explicit about the deviation, why it is there, and who it is for
      • Clearly signpost the transition back to the main story
      • Usually takes more words and more writing, which is something many authors struggle with

Problem 4. Not enough self (yet)

   Problem
      • Stodgy, overly cautious, formulaic text
      • Unclear connections
      • Weak or no message (often also repetitive)
      • Referencing poor/unclear (performative rather than informative)

   Solutions
      • Highlight where they need to say more
      • Tell them if you see bigger points that need to be made
      • Ask why they are referencing specific references, ask them to be explicit with the reader about this (show them how)
      • Make them write the cover letter!

Problem 5. Graphical ‘voice’ doesn’t fit field

      • There are field norms for graphical representations
        ─ meeting reader expectations enables faster comprehension
      • Can be good to upend norms in pursuit of new paradigms
        ─ sometimes just gets in the way of the message

The ChatGPT elephant in the room

In addition to these problems, we were given something else to confront: addressing the ChatGPT elephant. After analyzing the good, the bad, the beautiful, and the ugly characteristics of this new tool for generating text, Kate assessed the emerging relevance for its use in scientific writing. For those senior-level users who have already developed a personal voice and know what they need to develop their text, but who are also aware that the generated text needs careful editing, ChatGPT can be a useful tool. However, many junior-level authors are not confident enough of their own voice to make the appropriate corrections in a text generated by ChatGPT. There is a need for deeper understanding of two issues. First, teachers and editors need to help students to develop a critical eye early and teach them how to judge a text on what is expected and respected in their discourse community. We also need to help our students understand voice, how it is expressed, and why it is critical in engaging with their scientific storytelling and, consequently, with their readers.

Questions to ponder

Kate finished her presentation with two questions to ponder for the future:

• If a writer can ‘instantaneously’ merge their own writing with a field-average style, will the aspects of their writing that express their individual voice still be crucial for establishing their intellectual integrity?
• Is it only the data that matters, or is it also what you think of the data that matters?

On everyone’s behalf, I would like to thank Kate for an excellent presentation that clearly articulated the issue of recognizing the critical role of voice in scientific writing. For those of us teaching scientific writing, it is a call to go beyond this recognition and focus on the problems and their solutions facing junior-level researchers in navigating a personal voice in their storytelling.

Resources

Hyland, K. (2005). Stance and engagement: a model of interaction in academic discourse. Discourse Studies, 7(2), 173‒192. https://doi.org/10.1177/1461445605050365

Yasuda, S. (2022). Natural scientists’ perceptions of authorial voice in scientific writing: The influence of disciplinary expertise on revoicing processes. English for Specific Purposes, 67, 31‒45. https://doi.org/10.1016/j.esp.2022.03.001.