---
layout: post
title: "The politics of style: on styleguides in scholarly communications"
categories: [open access, style]
tags: [style, scholarly communications]
published: True

---
In the past couple of weeks I've had a number of emails about the styleguide for the [Open Library of Humanities](https://www.openlibhums.org). Queries range from "my discipline does things differently, is that OK?" to "why bother maintaining your own style guide, why not just refer to a named style like Chicago?" Both good points. My own thinking actually goes further, though.

Firstly, the most important thing that I'd like to flag up is that we have the following clause in our submission guidelines: "This journal recommends, but does not require, the Harvard system of references; see below for examples of how to format. Authors not submitting in Harvard style must ensure their references are internally consistent within the manuscript." When running a journal that spans many different disciplines, it is impracticable to enforce a single style of writing and reference. This is why we also say that "Research articles should be of an appropriate length for the discipline. While we recommend 8,000 words, this is flexible."

Secondly, however, I'm always led to a certain question that doesn't seem to be terribly popular: why don't we just do away with consistency between articles in a journal? So long as the article itself is internally consistent with its referencing and style of writing (and in accordance to a specified styleguide), then why do we try to homogenize throughout the container (the journal)? Given that semantically rich JATS-encoded bibliographies can be re-styled to whatever output format we want, the insistence on submission in a specific reference style is definitely nonsensical and just creates a lot more work for an author who may have to go through different rounds of submission to different journals.

Now, of course, I can think of several reasons why we have adopted this practice. Journals, in their hard copy forms, used to be read as coherent issues. I say "used to" because I would contend that, in the present digital age, this has disintegrated into disaggregated reading practices, with perhaps the only exception lying in the case of special issues. Certainly in my own research practice, I do not read journal issues cover-to-cover. Perhaps this is my loss but it is more and more common. And so I often genuinely have no sense of whether the journal enforces a coherent styleguide throughout.

Another reason is the need to "professionalize" a publication. Starting new journals, in particular, is hard because the activation energy to attract submissions is stacked against new publications. They don't yet have brand-name recognition. And so, if you want to avoid looking like some kind of scam journal, it's good practice to present a strong aesthetic coherence and an attention to detail. But we also know that the aesthetics of the piece are hardly what we should be valuing most highly in scholarly communications. It should, at the least, be a functional concern (good typesetting and well-written prose is easier to read), but should not, at the most, be some determinant of quality of a publication. If we do, perhaps we'd need some sort of awful Journal Aesthetics Factor (JAF) to equate stylistic coherence with quality.

My view on this, which remains under constant revision as I think about it more, as of July 2015 is:

1. We should dispense with the insistence on unity of style in the majority of cases for references and writing within journal containers. This is out of consideration for authors and out of pragmatics in trans-/multi- disciplinary spaces.
2. We should, however, insist that authors comply with a respected style and referencing guide of their choosing (and specify what this is) on a per-article basis. This is out of consideration for readers.
3. Journals should not write their own custom style guides. This is out of consideration to authors.
4. We should insist on the role of aesthetics in typesetting and production out of consideration for readers. This should not be seen as a proxy for quality or professionalism, though.

Then there are some exceptions and questions:

1. We may wish to continue insisting on unity of style when work is curated into special issues. Certainly one way of achieving this for references is through XSLT rendering of JATS element-citations in a unified style.
2. Do I feel differently about this for book chapters in edited collections? Probably, as per point #1 on special issues, above, which I consider to be akin to edited collections. The materiality of the codex object seems to have some bearing on my thought here.