The Rules and Norms of Philosophical Writing in the Analytical Tradition

What are the rules and norms of analytical philosophical writing? Here is a selection that I think many of us ‘insiders’ would recognize:

• • Use short, simple sentences.

• • Do not use elaborate (‘flowery’) language – choose simple vocabulary.

• • Always say at the outset what you're going to argue and, if possible, also indicate to the reader (in a general way) how you're going to argue it.

• • Use numbers to aid clarity of structure (e.g. ‘I shall consider three arguments for p; and then two objections to the first of these arguments).

• • Avoid saying anything about what you think or feel.

It may seem to those who have been taught such rules as these (among whom I include myself, of course) that they present natural and obvious principles by means of which to ensure maximum communicative clarity. But in fact, it is not difficult to cast doubt on the idea that any of them is really a crucial aid to the (laudable) aims of attaining lucidity and precision which are likely to be central to the analytical philosopher's understanding of how to perform their task well. In the next section of the article, I shall attempt to offer a line of argument which I think probably implies that there is a case for supposing that all of the rules I have just mentioned are gendered in certain ways. But quite independently of that possibility, one might wonder in any case just how obvious it really is, when one thinks harder about it, that these are self-evidently good rules for the writing of philosophy. Take the injunction to use short and simple sentences, for example. Is it really true that following this advice is an obvious aid to clarity? How clear one finds different styles of writing might differ from person to person, for a start – it depends on what one is accustomed to. And although it might be tempting to think there must be a presumption in favour of the thought that a short sentence is easier to understand than a long one, it needs to be recalled that what we are usually attempting to understand in philosophy is not single sentences but rather whole trains of thought – and once one bears this fact in mind, it becomes much less clear that prose entirely written in short, simple sentences must always have the advantage. Arguably, there are ways in which resolving to eschew the complex sentence might hamper rather than improve the clarity of what is being said, for what can sometimes be unclear, faced with a mere concatenation of short sentences, is what the relationship between them is supposed to be. Long sentences, containing more than one subclause, can do an array of things which simply cannot be done so readily by a concatenation of short sentences, for they are adept at displaying relation. Qualifications to – and clarifications of – a central claim, for example, are often most clearly done by adding a qualifying clause (e.g. ‘although’, ‘despite the fact that’, ‘on the assumption that’, ‘whether or not’) – and sometimes a sentence which includes more than one such qualification provides what is sometimes the best and clearest way to present the propositional relationships one wishes to convey. Hume is a particular master of use of the long sentence to convey these sorts of propositional relations, sometimes presenting an entire argument within the confines of one, as here:

Even I might be prepared to admit that Hume goes a little too far here in what he hopes to pack into the confines of a single sentence – but my point is that one can see in a sentence such as this how the skilful use of connectives (‘thus’; ‘and therefore’; ‘yet’, and so on) can make very plain in a way that is actually remarkably concise what it is that he intends to argue. In one fell swoop, he bats away an anticipated objection, makes plain the precise assumption on which his own conclusion rests, and makes his claim. It is doubtless true that such a long sentence as this takes longer for a reader to parse than a shorter one, but once parsed, the precise line of thought being offered by Hume here is beautifully and elegantly displayed. Hume is also a skilful wielder of that useful device, the semicolon, which can indicate less clunkily than an explicit statement that what comes after it is intended to be closely connected in some way (for example, as a justification of, or example of, or evidence for) what comes before. The semicolon has many detractors, but in my view, it represents a very useful means of indicating that one intends a relationship between the two parts of the sentence to be inferred. It is just not clear that one's task as a reader in reconstructing a line of thought is automatically easier when it involves the attempt to understand the intended relations between separate, shorter sentences. Obviously, there are limits to what can sensibly by done within the confines of a single sentence and there comes a point where the difficulty (for a reader) of parsing what has been said outweighs any advantage that can be gained by the explicit representation by the shape and structure of the sentence of the structure of certain intra-propositional relationships. But it is a balancing act – and writing in short sentences is not obviously better suited to the clear presentation of argumentation than a style which avails itself of the possibility of displaying the intended propositional relations by means of the rich array of connectives (and significant punctuation) which language has at its disposal. Both short and long sentences have their uses and a judicious mix of the two seems like the thing to recommend.

Vocabulary is another target of the analytical philosopher's rules. A preference for simpler language might seem to have a very obvious justification – plain-speaking has an obvious democratic appeal and perhaps there is no point reaching for a long word likely to be unfamiliar to one's readers, when a shorter and more straightforward one would be just as good. But first, it is rare that words really are perfectly intersubstitutable for one another (even leaving aside issues I shall come to later, such as the importance of word-choice to such things as the rhythm and balance of a sentence); and second, it must be noted that analytic philosophy's operation of its own edicts on this score has in any case always been very partial. An exception has always been made, for example, for such things as technical language, logical symbolism and terms of art. I do not dispute in the least the need for these resources in certain kinds of philosophical writing about certain kinds of topic; without these resources, indeed, certain points are pretty much unmakeable. But it can also be true that the more rarefied reaches of ordinary, non-technical language make expressive power (as well as colour, connotation, etc.) available to a writer that they would lack without those particular and perhaps less familiar words or phrases. What exactly would be the reason for not availing oneself of those resources as a writer, since they exist? It is, of course, obviously sensible always to write in a way likely to be accessible to one's readers – so in writing an A-Level textbook, for example, it is clearly a good idea to avoid words one's readers are not very likely to have come across before. But one must also avoid the dangers of patronization. No one can learn those less familiar words and the thoughts to which they might give access except by being invited by writers to share them.

Perhaps it will be said that I have grossly misrepresented the spirit in which the rules and norms of writing in analytical philosophy are usually offered. It might be claimed that they have only ever been proffered in an ‘other things being equal’ kind of spirit, that they have always been intended as guidelines rather than unbreakable rules. Moreover, it might be insisted that in any case, they are mainly supposed to be for students, that they are recognized to be ‘training wheels’ which can readily be discarded once a certain level of proficiency has been reached. But with respect to the first point, I think I have already given reasons for thinking that there may not exist even very persuasive ‘other things being equal’ style justifications for some of analytical philosophy's rules; and with respect to the second, I think there is evidence that the norms in question continue to be extensively obeyed by professional analytical philosophers (particularly in journal articles) and hence that the training wheels remain in place long after philosophers are deemed to be fully-fledged, independent academics. Patricia Nelson Limerick in a New York Times book review once compared academics in general to buzzards that have been wired to a branch for so long that even when the wire is pulled away, they continue to believe they cannot fly freely:

I think it may be that in analytical philosophy, we experience our own particular version of this chain of misinformation and misunderstanding, with each of us trying to second-guess ‘the other guy’ whom we imagine is in charge of making the decision about whether or not to publish our papers – or indeed trying, when in the role of referee, actually to be that other guy – the guy everyone is expecting us to be. But perhaps we thereby unwittingly encourage a rather stultifying conservatism in philosophical writing, of which we would do well to shake ourselves free.

Having tried to suggest, then, that at least some of the principles we are encouraged to obey are not as unquestionable as they are sometimes made to seem, I turn now to the second question in which I am interested: whether analytical philosophical style might in addition betray a certain kind of gender bias.

Gender and Style

Writing style is relevant to the position of women in analytical philosophy in at least two different ways – it bears on the appeal that written philosophical texts have to women; and it bears on the ease with which, as writers, they are able to adopt analytical philosophy's preferred stylistic principles. We might think, then, that answers to at least the following two questions are relevant to the present issue: Do women show a tendency to have different reading preferences from men? And do they show a tendency to write in ways which are different from the ways in which men write, in certain identifiable respects?

Both these questions have been quite extensively studied empirically by educationalists and linguists and the answer to both appears to be ‘yes’. Women read more than men overall, and they especially read more fiction than men (e.g. according to Weiner Reference Weiner2007, men accounted for only 20 per cent of the fiction market, according to surveys in the US, Canada and the UK). Moreover, and more importantly, the relation of boys and girls to literacy while they are growing up tends to be different. In childhood, boys are more inclined to focus on books which impart facts (books about e.g. space, football, dinosaurs, animals) and hence which tend to have a more ‘informational’ presentational style, whereas girls make more fictional selections (Smith Reference Smith2004). The influential educationalist Louise Rosenblatt theorized a distinction between ‘efferent’ and ‘aesthetic’ reading, the former done mainly for the purposes of information-acquisition and the latter characterized by a kind of immersion in which ‘the reader's attention is centered directly on what she is living through during her relationship with that particular text’ (Reference Rosenblatt1978: 25), and there is evidence that girls are more inclined towards the relevant kind of aesthetic immersion than boys. In addition, girls are more encouraged than boys, both at home and in schools, to write creatively and at length for pleasure (e.g. to write stories and plays), whereas boys often develop an early sense that both fiction-reading and creative writing are mainly feminine activities. Insofar as boys do eventually become successful writers, it has been hypothesized that they are incentivized at an early stage by the peer respect they gain from the knowledge they acquire in areas which are of interest to many boys – and in this way develop a positive correlation between the reading of these informational materials and masculinity (Smith Reference Smith2004). Reading thus becomes a source of affirmation for these boys – and the facility in writing which comes from the more extensive reading they are incentivized to do then develops. Girls, on the other hand, are more likely to develop an early affinity for, and pleasure in, language itself and are more likely to exploit its expressive powers in the writing of letters, diaries and poems from a young age (National Literacy Trust 2016).

These differences in developmental experiences appear to feed through in various ways into the writing of men and women as mature adults. Much controversy surrounds what precisely are the features that are associated with men's and women's writing style, but the fact is that automated text categorization programs trained on a wide selection of written materials are able (while controlling for genre) to infer the gender of the author of a formal written document with approximately 80% accuracy (Koppel, Argamon and Shimoni Reference Koppel, Argamon and Shimoni2002) – and that shows very clearly that important differences in writing style must exist. It has proved quite hard to characterize some of these differences at a level which makes it easy to understand and interpret them. Some of the ‘ground floor’ differences which are assessable by computational means are that women are much bigger users of personal pronouns than men; whereas men use many more so-called ‘determiners’ – words like ‘the’, ‘a’, ‘this’, ‘every’, ‘many’, ‘two’, etc., which occur in conjunction with noun-phrases. Numbers, in particular, are used far more often by men than by women in writing (a fact which is obviously relevant to another of analytic philosophy's ‘rules’ which I mentioned above). It has been hypothesized that the greater use of determiners by men and of pronouns by women is correlated with a difference in general approach to the presentation of information (Koppel, Argamon and Shimon Reference Koppel, Argamon and Shimoni2002). Linguists have used the terms ‘involved’ vs ‘informative’ (Biber Reference Biber1995) to characterize some of the stylistic differences which seem to be involved – women are more likely to explain what things they are talking about by highlighting relationships to people (e.g. ‘my point’; ‘her desire to avoid hostility’, ‘his predicament’); while men are much more likely to adopt impersonal turns of phrase (e.g. ‘the first point’; ‘the desire to avoid hostility’; ‘a disposition to seek approval’). Women are also more likely to write in ways which forefront their role as writer and their imagined reader's role as reader (which generates an increased use of first and second person pronouns), rather than seeking to keep these in the background. We need not here wade into the controversial arguments which surround the question whether these gender differences in women's and men's relationships to reading and writing are ultimately rooted mainly in biological or in societal factors. For present purposes, we need only note that research seems to give good grounds for supposing that women and men do indeed have different relationships both to reading and to writing and that a limited consensus is beginning to emerge about what the nature of some of those differences might be.

How might such differences be relevant to the way in which men and women respond to analytic philosophy as readers and writers? Proper research on the matter would be needed before one would be able to draw any firm conclusions, but here are a range of hypotheses that strike me as fairly plausible, given what we know about the developmental, educational and linguistic facts:

1. (i) Men would be more likely than women to show a reading preference for the kind of lean, concise and ‘unflowery’ prose which analytical philosophy encourages because of their greater levels of exposure to the leaner, concise and unflowery prose more typical of the informational reading in which they preferentially engage.

2. (ii) Men would be more likely than women to feel comfortable writing in the lean, concise manner recommended by analytical philosophy's rules and norms.

3. (iii) Women would be more likely than men to wish to import into their philosophical writing some of the methods and techniques typical of fiction – and perhaps to feel stymied when they discover that these methods and techniques are not generally encouraged. Obvious examples of such techniques would be the use of imagery, of word choice that is influenced not merely by semantic properties but also by colour, rhythmic features, etc. They might also be more disposed to employ novelistic techniques such as suspense which are frowned on in philosophy (e.g. by leaving the solution to the apparent philosophical problem outlined at the beginning of the paper to be revealed at the end, thus making the reader wait for a dénouement – rather than up-front, as standardly recommended).

4. (iv) Women would be more likely than men to include themselves in their writing – to say e.g. ‘I think’ or ‘it seems to me’, or ‘I have the sense that’ and to write in ‘involved’ ways, rather than in the dispassionate style recommended by our writing rules.

Here is not the place to try to defend any of these hypotheses. I merely note that they may be arrived at by fairly plausible varieties of inference from differential gender tendencies for which we do appear to have some evidence; and, moreover, that if any of them is correct, then there might be a case for thinking that the analytic style of writing in philosophy is gendered.

I want to end by asking you to consider these ten rules for writing, which were offered by Nietzsche to the Russian intellectual Lou Andreas-Salomé, in a series of letters, and collected under the heading ‘Toward the Teaching of Style’:

1. 1. Of prime necessity is life: a style should live.

2. 2. Style should be suited to the specific person with whom you wish to communicate. (The law of mutual relation.)

3. 3. First, one must determine precisely ‘what-and-what do I wish to say and present,’ before you may write. Writing must be mimicry.

4. 4. Since the writer lacks many of the speaker's means, he must in general have for his model a very expressive kind of presentation of necessity, the written copy will appear much paler.

5. 5. The richness of life reveals itself through a richness of gestures. One must learn to feel everything – the length and retarding of sentences, interpunctuations, the choice of words, the pausing, the sequence of arguments – like gestures.

6. 6. Be careful with periods! Only those people who also have long duration of breath while speaking are entitled to periods. With most people, the period is a matter of affectation.

7. 7. Style ought to prove that one believes in an idea; not only that one thinks it but also feels it.

8. 8. The more abstract a truth which one wishes to teach, the more one must first entice the senses.

9. 9. Strategy on the part of the good writer of prose consists of choosing his means for stepping close to poetry but never stepping into it.

10. 10. It is not good manners or clever to deprive one's reader of the most obvious objections. It is very good manners and very clever to leave it to one's reader alone to pronounce the ultimate quintessence of our wisdom. (Nietzsche, quoted in Salomé Reference Salomé and Mandel2001/1894: 77–8)

It is hard not to be struck by the extreme contrast between this set of rules and the norms under which we currently operate in analytic philosophy. Our contemporary rules say nothing about ‘life’, about ‘expressiveness’, about the felt, rhythmic qualities of language which Nietzsche compares to gesture, about the need to entice the senses. But perhaps it is time we thought more about such things.

Why do I think we should think more about them? To be clear: it is not primarily because I think the style in which analytical philosophy is written is a major contributor to the underrepresentation of women in the field. Differential preferences for certain writing styles seem unlikely to play more than a fairly minor role in explaining the differential attraction of the discipline for women and men. As Louise Antony (Reference Antony2012) has argued, it is likely that a wide range of factors are at play in the underrepresentation of women in philosophy which come together in a peculiarly marked fashion in the discipline (what she refers to in the title of her article as a ‘perfect storm’ of interacting factors, including stereotype threat; gender schemas; differential assignment and uptake of ‘service work’ by men and women; implicit bias, etc.) – and writing style seems unlikely to be among the most important of these multiple contributing factors, given the obviousness of the great importance of some of the others. No, I want us to think more about the rules because I think they are wrong. If we are going to have disciplinary writing rules, we need to be surer than we currently have a right to be that we can differentiate between those which are genuinely connected with the laudable aims (on which philosophers pride themselves) of clarity, accuracy, measuredness and objectivity; and those which have rather become historically entrenched for reasons that are not really directly connected with these aims at all. The point is not so much that the rules we have betray the personal and gender-inflected preferences of those who have mainly been the guardians of our discipline – although there may be some truth in that. It is rather that, as Bernard Williams (Reference Williams, Bunnin and Tsui-James1996) has pointed out, the identity of analytical philosophy is tightly bound up with its ambition to be a discipline that can sustain ongoing, cumulative research which adds to the sum total of human knowledge; and which can thereby clearly differentiate itself from the sort of philosophy undertaken by those like Nietzsche, who was not primarily concerned with whether he had shown or proven anything. Our writing rules therefore reflect the ambition to be as unlike (say) Nietzsche (and those other suspicious ‘continental’ philosophers) as possible and as much like natural science as philosophy can ever manage to be. Williams did not think these claims by analytical philosophy to be able to add to knowledge were empty, and neither do I. I think there is progress in our discipline and I think we find things out. But we don't need to feel that that implies that we cannot avail ourselves of techniques, modes of expression, appeals to the senses, etc. in order to align ourselves with properly philosophical aims. I hope that reflection on the possible genderedness of the rules we currently operate might encourage the sense that we could well afford to be more adventurous and imaginative as writers of philosophy than we mostly currently are.