author : Steven Pinker

Do we generate our utterances and recognise those of others by using some deep set of rules? Or do we have a neural network implementing an associative memory look-up? Some linguists argue for one, some for the other. Pinker, in this lucid and fascinating account, argues for both.

I never knew there was so much to regular and irregular English verbs! This is the main example Pinker uses to discover and explain the existence of both mechanisms in our heads. His secondary example is regular and irregular English plurals, and he touches on a wide range of examples from other languages and other parts of speech. His argument is that we use a kind of associative memory dictionary to generate distinct words, including the irregular cases (the Words part of the title), and a rule based system to generate the regular, or default, cases from word stems (the Rules part).

All this is illustrated by a fascinating array of marvelous examples of common speech acts and speech errors, starting from the observation that young children go through an “overgeneralisation” stage where they say things like bringed and putted instead of brought and put, on to more complicated cases like just why the past tense of to ring varies between “ringed the city” and “rang the bell”, backed up with statistical evidence from careful experiments. As always in this kind of study, there is an examination of paired brain problems, here Alzheimer’s (affecting mostly Words) versus Parkinson’s (affecting mostly Rules), and Specific Language Impairment (high intelligence but impaired language skills) versus Williams Syndrome (retarded intelligence but excellent language skills).

The whole treatment is covered in great depth, yet is eminently readable, because it is accompanied by illustrative anecdotes, cartoons, historical insights (why Shakespeare wrote that apostrophe in “star cross’d lovers”, and why the events of 1066 means English is so more regular than the closely related German), fascinating little details from other languages (French, Hungarian, Arabic, Hebrew, Chinese, even the “Stone Age” Arapesh), and a light, humorous style (which other writer would illustrate the first person present tense of “to be” with the example sentence “I am the walrus”?). Fascinating, insightful, readable, and convincing.

I’m falling behind – the rest of the world is writing good stuff faster than I can read it. So, rather than trying to go through authors’ back catalogues in order, I’m deciding to skip missed books, and just go to the latest one. Which is a shame in Pinker’s case, because he writes beautifully, and has deep interesting important things to say. But, world enough and time.

Here he is investigating the concepts and structures with which we think, in a series of chapters that look at different aspects of the way we use language. So, for example, he looks at certain apparently irregular sentence structures that children learn with little difficulty, even though they can’t possibly have heard all the examples, and shows they can be explained by a relatively simple set of underlying concepts of time, substance, having, moving, causing, etc. He looks at metaphor, and the strong claims of Lakoff et al that we think almost exclusively in metaphors. He shows it is more complicated than that. He looks at the way we name things, and how the thing we name is still that thing, even if everything we know about it has changed. He looks at the taboo vocabulary, grammar, and semantics of swearing (including some of the linguistic delights I first encountered elsewhere), and shows that it is more complicated than simple explanations of trying to shock, or to let off steam. He looks at indirect “polite” speech, and again shows it’s more complicated than simply communicating directly, or saving “face”.

All this is told in a lovely lucid style, liberally peppered with examples, quotations, and cartoons. Amusingly, it has the “which of these shapes is called malooma, which takata?” question that appeared in the immediately previous book I read: here in the context of underlying thought concepts, there in the context of synaesthesia. Everybody gives the same answer: how, given the words are just made up?

The man is clearly drunk on words, be it comprehensive lists of verbs or nouns in various classes, or wonderful turns of phrase in his prose. Just consider the delightful “the Pluto formerly known as a planet”. The Stuff of Thought is great stuff.

Steven Pinker has written several popular science books in his areas of expertise: language and cognition. These books deliver profound insights, and they are also tremendously readable, delivering those insights with style and verve. In this latest offering, he moves from presenting research results in a readable manner, to presenting advice on how to write readably, fittingly also in a readable manner.

Writing well is non-trivial. Pinker describes his own process.

The book is divided into six meaty chapters, each capturing a different aspect of good writing. This is not about picky little examples, held up by those anxious grammar police as the epitome of style; rather it covers the deeper structure and content of prose. This means that simple rules of thumb, such as "avoid the passive voice", should not be used indiscriminately: although the passive often has the effect of moving attention away from the guilty agent (such as yourself), sometimes it is needed to focus attention onto the important agent.

First, Pinker introduces good writing in general, dissecting examples of good, and bad, prose, pointing out where they work, and where they fall apart. I did not always spot which were the poor examples until that subsequent dissection made it clear; I am perhaps too used to reading inelegantly written text for it to sound out of tune to me.

Next, Pinker discusses classic style, a particular style for writing clear, compelling prose, eschewing obfuscation. He summarises it thus:

That passage itself, along with most of the book, is written in classic style. Pinker then takes many examples of convoluted, turgid academic prose, and shows how to rewrite them in a clearer, comprehensible, livelier style.

In chapter 3, Pinker covers The Curse of Knowledge: the writer knows a lot more about the subject matter than does the typical reader, and can bamboozle them if they are not careful. It is hard to get the right level, between confusing and patronising the audience. Pinker provides a few tips.

Next we get a chapter on an area of Pinker’s expertise: grammar. His aim is to show how an understanding of grammar can not merely make sentences grammatical, it can also help prevent grammatical ones from being difficult to parse and potentially ambiguous. From this chapter it is clear that English grammar has changed from what I was briefly taught in school many years ago. Then it was all nouns and verbs, sentence subjects and objects; now it appears that there are different categorisations, and finer distinctions:

There are lots of good examples in this chapter, and Pinker uses grammatical theory to demonstrate why they are problematic, and how use of grammatical structure can improve them. With his usual lightness of touch, Pinker distinguishes ways to advertise a pair of panel discussions:

1. 1. A panel with four professors on sex
   2. A panel on sex with four professors
2. 1. A panel with four professors on drugs
   2. A panel on drugs with four professors

In the first case, formulation I avoids a potential mis-parse; however, in the second case, formulation II is the unambiguous one. There is no simple rote solution to be employed. This is the toughest chapter in the book, and Pinker occasionally falls prey to the Curse of Knowledge himself here.

In the next chapter, Arcs of Coherence, Pinker moves up from discussing single sentences to addressing the overall structure of a piece of prose. This includes advice about stating the topic early on, to give the reader something to hang the rest of the text on, and then to present the rest of the text in a logical order that makes sense to the reader. These potential platitudes are enlivened through a great choice of examples.

The final chapter reverts to grammar police style: rights and wrongs. Pinker subverts the usual prescriptive style, however, taking time to explain why most of the grammar police edicts are flat out wrong. But at the end, even he cannot resist his own list of preferred and problematic usages.

This is an excellent guide to clear writing, and I would recommend it, along with Williams’ Style, to all aspiring communicators.