Rethinking Writing, a’ la Roy Harris (I)

Writing does not represent language, it is language (meanings are not independent of their codes/signs)

Roy Harris’ (1931 – 2015) oeuvre is vast, and I am just skimming its surface. He was Emeritus Professor of General Linguistics at Oxford and the following reflections are based on his ‘Integrationist Notes and Papers 2009-2011’ and ‘Rethinking Writing’ (2001).

In a nutshell, Harris argues that what we mean by language resides in the way we actually communicate (he’s a descriptivist). Language  is not something extraneous to how it is used (he’s not a prescriptivist). This means there is no ideal, prescriptive way of understanding language because language is a system of signs that integrates human behaviour (social dimension).

He is known for his historical accounts of writing, his views on literacy, and his semiological (sign theory) critiques of:

1. De Saussure‘s structural linguistics: Harris accused De Saussure of being incompatible with his very own theory, mainly because his structuralism could not simultaneously reconcile the autonomy of the linguistic system with an internal logic that necessitated validation from an outside world (2011, p, 51);

2. the idealised linguistic system of the early Chomsky: for Harris, Chomsky’s segregationism is redundant from the outset because there is no such thing as an ‘idealised speaker’; language does not exist independently of the speaker (2011, p.12);

(both De Saussure and Chomsky are ‘mentalists’ (cognitivists, in Chomsky’s case)

3. Wittgenstein‘s theory of language games in the Philosophical Investigations: for Harris, it makes no sense to assume that language can ever be complete in the sense that Wittgenstein assumed, because this historical/organic approach to language would entail that what we speak/write NOW is not ‘complete’ (i.e. it assumes that for a language system to be complete, it would have to reside outside of what we do NOW when we speak/write, which is why no dictionary, grammar, etc. can ever be complete) (2011, p.20);

4. Bloomfield‘s (and Skinner’s) behaviourism: Harris claims that behavioursists paid no attention to context and that their linguistic theory reified ‘behaviour’ to being a mere ‘residue’ of linguistic evidence (2011, p.68).

Harris’ legacy and contribution to the field of linguistics – born of these critiques – is known as integrationism, a theory which boldly posits that the object of linguistics is not language, but context (Harris, 2011: 1):

… linguistics stands as much chance as phrenology of passing muster as a science of language (Harris 2011, p. 58)

In other words, ‘linguistics’ is a pseudo-science.

Language-learning does not proceed by a strategy which requires the learner to ignore context and concentrate solely on linguistic residue. On the contrary, in order to progress, the learner must grasp the contextual implications of everything that speech and writing involve (Ibid, p. 68)

In other words, a skills-based / transmission model does not work when it comes to meaning (cf. Hogan 2014)

For the integrationist, the ways we use words are open-ended practices, constantly modified in the course of interpersonal communication (ibid)

In other words, the ‘golden standard‘ of the way we use words can be constantly re-negotiated. This has implications for what I mean by ‘academic writing’.

Integrationism rejects the attempt to draw any clear dividing line between linguistic and non-linguistic activities (ibid)

In other words: multimodality (Lillis 2013, Chapter 2).

What might Harris mean by all of this, and is it a fruitful theoretical lens for my own research on how academic properties/aka academic meanings may be emerging from their written forms (btw – this post is a pure example of writing as process, as a way to clarify concepts for the writer (me) in a way that speaking would not be possible (Emig, 1977: 123; Harris, 2000: xi).

What is literacy?

Literacy* involves the creation of a social status for initiates, and an acceptance of the scale of values over and above those which the task entails (Harris, 2001: 6)

*Literacy, here, is to be understood as ‘writing’.

This quote presupposes that the worth of a ‘civilisation’ is to be judged by its literary output understood as its writing (Harris 2001, pages 8 and 9). According to Harris, though, this is a circular argument because once the powers that be have decided that literacy is going to be the hallmark of civility, then it follows that when you ask others what marks out a ‘civilised society’, the answer is clearly going to be: ‘being literate’.

It also implies that becoming literate is not the result of mastering a set of skills, but of accepting a ‘scale of values’. Once those values have been accepted by the letterati, then the letterati themselves define literacy for everybody else (neat!).

Reflection: This allows me to question who defines ‘literate’ in today’s society? Who defines ‘academic’?

Harris argues that being ‘literate’ is not the exclusive domain of ‘writers’ and that “poetry, narratives, and other verbal compositions that engage the intelligence and the imagination” (Harris 2001, page 11) are also forms of literature (in this sense, he is opposed to Ong, p. 13).

Harris is adamant that writing does not depict/represent speech, or language. His arguments for this are many, including the following:

  • an alphabet cannot capture all the nuances of speech (2000, p. 18); as such, writing cannot be a surrogate of speech;
  • “writing is the lifeless shell or trace that language leaves behind” (ibid) (Harris invokes Plato, here ….. hmmmm….but this assumes that language exists independently of writing, which seems to contradict Harris’ own theory? (check);
  • writing is not the same as its “material residuum”; if it were, it would be like saying that a portrait IS who it depicts;
  • speech is not divorced from the speaker; writing is not divorced from the writer (Plato).

Here is why Harris is an integrationist:

… languages are not conceived as systems independent of their use in communication’ (quoting Matthews 2007, p. 197, in Harris 2011, p, 23).

Language is not an ‘operation’ (of the mind) at all, but a faculty possessed by creatures who engage in verbal communication (2011, p. 60)

If anybody out there has any views on this, I would be glad to hear from you.


Harris, R. (2001) Rethinking Writing, Continuum Press

Harris, R. (2011) Integrationist Notes and Papers 2009-201, ISBN 978-07552-1341-2

Lillis, T. (2013) The Sociolinguistics of Writing, Edinburgh University Press

Pádraig Hogan (2014) ‘Recovering the lost métier of philosophy of education? Reflections on educational thought, policy and practice in the UK and farther afield’. JOURNAL OF PHILOSOPHY OF EDUCATION, 48