shytone  books  music  essays  home  exploratories  new this month

book reviews



Gunther Kress: Learning to Write (2nd ed.)
(Routledge: 1994)


“Speech is learned before writing. Hence the grammatical and textual (and of course the phonological) rules of speech are learned first, and form the basis of the child’s knowledge and use of language. Speech is characterized by the immediate presence of an addressee (a single or multi-person addressee). The audience tends to be known to the speaker...spoken texts are typically created in interaction...[and] much knowledge may remain implicit.... What are the effects and implications of these features for children’s writing? ...Whereas in speech, the child creates a text in interaction, now he or she is, for the first time, forced to construct a text without the guide, the prodding, the stimulus of the interaction. Producing speech under such circumstances is very much an abstract exercise.... [Moreover,] where the child imagines some addressee, the child may well leave as much information implicit as he or she would have done in the presence of this imagined addressee. Hence children’s early writing is frequently extraordinarily implicit, to the point of seeming disjointed.... [And] the structure of writing is fundamentally distinct from that of speech.... The development of the adult concept of the sentence...demands the development of planning, deciding which is to be the main clause (the main idea?) and how subsidiary clauses and ideas are to be integrated with the main clause. Conceptual and syntactic complexity here go hand in hand.”
(Kress, pp.35-7)

Given the undoubted importance of literacy to our society - as well as the many forms of evidence suggesting significant cognitive differences between oral and literate societies - it is rather troubling to note just how few books genuinely attempt to address the means by which we become literate - as if such processes were vaguely shameful, in some unspecified way... And as a resolutely non-Chomskian linguist - whose concerns center on the social and the developmental - Gunther Kress is very well positioned to address such questions, and does so insightfully - albeit with little concern for the requirements of the lay reader. And in doing so, he also provides crucial links between the cultural/psychological approaches of Bakhtin & Vygotsky and the oral/literate divide historical work of researchers such as Eric Havelock in the crucial realm of linguistics - a service for which he deserves a much wider readership.

For, make no mistake, this important subject is currently relegated to the neglected fields of pedagogy, being basically ignored by the linguistic mainstream, and not generally a major focus of developmental work. And Kress has a very good idea why this happens be so:


“One startling fact, and one that needs explanation, is the massive discrepancy between the amount of work which has been done on reading, compared to the work done on writing.... [This may be connected to] the types of linguistic theories which were current over the last fifty years. Theories which focus on structures below the sentence, on decontextualized sentences, on meaning as inherent in the individual linguistic item, on reading as a decoding skill, are unlikely to treat the learning of writing as problematic, nor indeed are they likely to be able to deal adequately with the textual aspects of writing.... Perhaps the very metaphor of ‘coding’ has proven to be the major problem...[as] from this point of view, decoding is the problematic part of the process: the encoder knows both meaning and code, the decoder knows only the code. [Moreover,] the ‘code’ metaphor has another effect in that it predisposes the user of the metaphor to regard the code itself as empty or vacuous, without meaning in itself.... It is true that the dots and dashes of morse-code contribute no meaning in themselves, [but] the same is not true of language.... [Furthermore,] the assumption in modern linguistic studies has been that there is a single language system, and its rules have been drawn out by an unacknowledged use of the forms of the written language. No doubt this sleight of hand (unrecognized by linguists themselves) was facilitated by the socially higher valuations of writing, which placed the normal forms of speech in the categories of substandard, incorrect, lower-class language.”
(Kress, pp.3-19)

Of course, there are always dissenters from any mainstream, but Kress belongs to a particularly important tradition in this area, which has insisted upon studying speech as such - rather than abstracted “examples” - and its key findings deserve to be general knowledge amongst the well-educated...instead of their current fate, which is to be ignored by almost all. Let’s hear from Kress, now, on the (major) linguistic divide between speech and writing, and the (strong) functional rationale which maintains it - matters which remain heresy to the vast majority of linguists:


“The...structuring of speech and that of writing proceed from two distinctly different starting points. The structure of speech starts from the question: ‘What can I assume is common and shared knowledge for my addressee and myself?’ This question, and its answer, are at the basis of the structure of speech. Writing starts with the question: ‘What is most important, topically, to me, in this sentence which I am about to write? This question, and its answer, are at the basis of the structure of writing. The cohesive and continuous development of a topic is thus paramount in writing, while the construction of a world of shared meaning is paramount in speaking.”
(Kress, pp.27-8)

“The sentence is not a unit of typical spoken language. The sentence belongs to writing, forming there the basic unit of textual structures. The sentence may occur in speech, as a borrowing from the syntax of writing, but speech, typically, is organized on the basis of clausal complexes which are not sentences. They may be long chains of clauses linked by co-ordination, or simply by being adjoined. While the sentence typically is a structure of main and subordinated and embedded clauses, the clausal structure complex is typically an aggregate rather than a syntactic structure. The thematic structuring of sentences [also] differs markedly from that of clausal complexes. Typically, each sentence is a construct with an internal structure which marks the thematic element of each sentence from the non-thematic. The treatment and development of topical material within the sentence is hierarchical and integrative. Within a clausal complex, the thematic structure is replaced by two structuring devices. On the one hand, there is a sequential development of topics, so that clauses in sequence may take over the theme/rheme structure within the sentence. On the other hand, superimposed on this structure is another structure, carried or expressed by intonation, which marks some elements... [to] the hearer as presenting some material as being already known to him, and other material as new to him.”
(Kress, pp.7-8)

“A hearer cannot usually pause, or check back over the message to make sure he or she has understood it. The structure of speech makes major allowances for this factor, by structuring information, [and] providing sound cues which highlight the relevant informational structuring...[which] is expressed through intonation. The speaker uses intonation to bracket together segments of his utterance which he regards as constituting one relevant parcel or unit of information...[and these are] marked off in speech by a single unified intonation contour and, above all, by containing one major pitch-movement.... The information units are motivated directly by the interaction between the speaker and the other participant...[and] each unit has an internal, two term structure, of ‘known’ and ‘unknown’ information... The greatest and most pronounced pitch-movement occurs at the beginning of the unknown segment...[which] tends to occur at the end of the information unit.... The two-part structure implies that there is always some common ground, some known information, between speaker and hearer. The unknown must be said, the known may be said, or it may be left unsaid and implicit.”
(Kress, pp.23-5)

“The structure and meaning of clausal connections in speech are highly articulated [by intonation] and capable of the finest nuance. In writing, this structure is mirrored to some extent in the system of punctuation, which is not, however, capable of expressing the same detail and precision.... Topic development in speech, within this clausal structure, is by sequence, restatement, elaboration, and intonational articulation. The evidence of ‘thinking on your feet’ is everywhere evident in speech. This is in direct contrast to writing, where there are, typically, no traces of immediate thinking. Writing is the domain of circumspection, of (self-) censorship, reworking, editing. The development of the topic is by another order: not by sequence, but by hierarchy. That which is more important is given structural prominence, the less important is structurally subordinated. Consequently, writing is the domain of a more complex syntax, typified by the sentence, by subordination and embedding of different types, by syntactic and conceptual integration. Speech is typified by the syntax of sequence, of the clausal chain, of addition and accretion.... There are other characteristics, too. The spoken text is longer, to make allowances for the different mode of reception, it shows repetition, allowing the hearer time to assimilate information. It also contains many features of an interpersonal kind, referring to the inter-relation of speaker and audience.”
(Kress, pp29-33)


As you can see - and, as I warned you in advance - Kress is definitely writing for a specialist audience, however important his arguments may be for the rest of us. Because we’re still - even most language specialists - under the sway of what Kress terms “folk linguistics”...in which sentences are characteristic of all well-formed expression, having never been shown the evidence as to how we actually (usually) speak. In particular, this knowledge should be central to the methods used to teach writing - at any level - for we badly underestimate exactly what this task involves...


“Learning to write has some of the features of learning a second language, including the initial ‘interference’ from the first language...[whilst] the co-ordinating, ‘chaining’ syntax of speech presents conceptual materials in a distinctly different form from the subordinating, embedding syntax of writing. The one points towards the order of sequence, the other points towards the order of hierarchy, [and] the habitual and unreflecting use of either may lead to differing modes of cognition.”
(Kress, pp.8-11)

If anything, this still underestimates the problems, since young children learn a second (or third) language easily by frequent immersion - whereas we have little or no genetic predisposition for the written word...however valuable we may find it, and the cognitive habits that come in its train. And, speaking of said habits, Kress’ detailed analyses of young children’s writings open up the mechanics of both oral and literate ways of thought, in a way which substantially enriches our understandings of just how these function.

It’s just a pity Eric Havelock isn’t still around to read these:


“Topic development, and the textual structures appropriate to it are more or less absent in the very early stages in the learning of writing. Some exceptions are provided by narrative writing, where the sequence of the events suggests its own sequential, linear structure to the child.... Narrative writing may thus have a specially important place in the learning of writing, in that it permits the child to develop textual structures and devices in writing by drawing on the child’s already established abilities in spoken language.... At this age...the child decides what goes together textually, and the sentence has to fit around it. [In contrast,] the literate adult knows what a sentence ought to be, and in writing this concept of sentence determines what goes into a sentence, and ideas have to be fitted to suit the form.”
(Kress, pp.60-76)

“The child’s early sentences suggest that they have a similar function in overall topic development to the paragraph in adult writing, and a similar internal function and structure to the adult paragraph.... It seems that one of the consequences of the achievement of the paragraph is that it frees the sentence from paragraph-like functions, thereby permitting the development of the sentence as a linguistic unit with structural, textual, and semantic functions in its own right. In the main, the effect is to reduce the content of the sentence. Because the paragraph has emerged as the unit which contains topically connected material, the sentence can now become the unit which contains one topically discrete component. [And,] as the child attains greater mastery over written syntax, much textual complexity and diversity is the subject of syntactic and topical processing, which gives rise to apparently simple structures...due to the prior close integration of material.... The effect on the sentence is to permit greater precision, integration, reduction and compression of content, and greater complexity, both cognitive and linguistic.... The development of this textual structure both permits and demands new modes of cognitive and conceptual organization. In this process, the receding audience or addressee is highly significant too.... The freeing from the specific addressee brings with it the possibility of a freeing from the dependence on the [immediate] order and logic of the real world [and,] as the addressee recedes, so the demands (however adequately or inadequately grasped) of the subject matter become foregrounded. The two processes together lead to formality [and] impersonality, but permit the development of abstract cognitive conceptual orders, expressed in the textual structures made possible and available in writing.”
(Kress, pp.94-7)

Taking the synthesis game a little further, it’s also very interesting to relate the next stages of Kress’ argument - where he takes a position on genre which is somewhat indebted to Mikhail Bakhtin - and juxtapose it with Mary Douglas’ approach to institutions as social cognitive models, and Richard Lanham on rhetoric as a basis for cross-disciplinary study. In all of these cases, I think, we can see a resolutely grounded (and usefully convergent) approach to  the key questions of social cognition, currently dominated by the windy speculations of postmodernism, in which “practice” is something to be talked to death, rather than closely studied.

Me, I prefer people who know enough to get their hands dirty...


“Learning to write involves the learning of new syntax, and of the new syntactic unit, the sentence. It also involves learning of the larger structures within which sentences occur...[and] textual structures have a world-ordering, even a world-creating function which is at least as significant as the world-encoding [?]  function of the sentence.... The child has to gain mastery over the forms and the possibilities of the different generic types, as part of the process of learning to write. The different genres each make their own demands in terms of their formal structures, their ordering of thematic material, their conception of knowledge, [and] these demands have their effects on the syntax at sentence and below-sentence level.... Inevitably, this learning takes place in the school...but if the teaching of writing has received little attention, the teaching of genre has had even less, with the possible exception of some literary genres.... [And so,] the major genres which the school teaches are taught by the way, implicitly: descriptive, scientific, technical, historical writing. This is true also of creative writing: children are asked to write ‘stories’ - with whatever motivation or purpose - without being given any teaching about the appropriate type of genre, or its main structural features. Consequently, children pick up the requirements of the different genres by osmosis, as it were.”
(Kress, pp.98-100)

“When children first start writing, they are quite unselfconscious of their role as narrator. So, while the surface of the text has signs of the narrator everywhere, from the child writer’s point of view there is no narrator present. [However,] when children have learned to write, they are fully conscious of their role as narrator. So while the surface of the text may show no signs of the narrator, from the child’s point of view, the narrator is everywhere present. In the same way, child writers may be quite unaware of any audience when they first write: the needs of a specific audience are not an issue, and hence are not met. [But] when children have learned to write...the audience may not be addressed overtly, but their presence is everywhere felt in the text; it is constructed for and around the perceived needs and demands of the audience. On the surface, the starting point and the finishing point have much in common; underneath, the two are clear, neat and complete inversions of each others.”
(Kress, p.124)

The further we progress through Learning to Write, the more it is evident that Kress is unafraid to tackle issues that would usually belong rather to developmental psychology - he is clearly (and usefully) familiar w/both Piaget & Vygotsky - and that the linguistic approach he brings to these offers valuable sidelights to some of the traditional questions in this area:


“The child does not make a systematic distinction between temporal and spatial sequence or succession; hence the child may be relying on one mode of order where the adult has several. The child’s logic may therefore be quite unlike that of the adult, bound more to the order of events in the world, and not able to abstract one kind of order from another.... [And,] in the same way that a concept such as the sentence becomes problematic for the child when he or she learns to write, so notions of order, causality, become problematic when the child is asked to ‘write a story’. In learning to speak, the child has learned some of the structures and contents of adult language unreflectingly...in interaction, often with adults. The interaction itself provides both the stimulus and the constraints for language use. In such a context, a child may learn syntactic forms as ready-made and unanalyzed units. In learning to write, this interactive motor for language use is absent, so that language has to be produced from scratch, as it were. Furthermore, linguistic structures present themselves in an objective and therefore problematic fashion - and the child responds to the possibility and opportunity of exercising choice in the selection of linguistic form.”
(Kress, pp.137-51)

“We need to distinguish between two theories of causality: the ‘powers’ theory of causality, and the ‘regularity’ theory of causality. The powers view appears, in English, as the commonsense view. It embodies the idea that...a causal connection is a real, active process, by which one thing produces another.... The regularity view, on the other hand, sees causal processes as merely the regular succession of states of affairs...[and] does not recognize the existence of actual causal powers.... While the powers view tends, in English, to be reflected by adult grammar, the regularity view tends to be expressed more often in the grammar of children...[as] the chaining syntax of speech tends towards the expression of the regularity view [whilst] the embedding and subordinating syntax of writing tends towards the expression of the powers view.... It is important to note that this is a tendency; adult language will frequently be found to show the syntax of the regularity view, in particular whenever the syntax of speech asserts itself, or where causal relations in the sense of the powers view are not understood or perceived, and therefore cannot be, or are not, either claimed or expressed. Interestingly enough, much scientific writing shows this tendency...[and] it would be a grave mistake to consider the regularity view naive, childish, immature, or anything similar. Indeed, the body of philosophical opinion since Hume would probably take the opposite position, if a choice were to be made.”
(Kress, pp.162-3)

“Our analysis...suggests that the child’s grammar can be consistently interpreted as a genuine alternative to, and not merely an underdeveloped form of, adult grammar...because this grammar is adequate to their cognitive and social needs at that time. To teach children adult grammar, therefore, a justifiable teaching strategy would need to be centrally concerned with broadening their experience...of being active causal agents, or manipulating real things and seeing their acts as the generative causes of real-world situations. Without such experience, children must see adult grammar as not only foreign to their world but irrational, since it makes claims about actions and forces when only events in sequence are experienced. With such experience, they will see the sense in adult grammar, or - if the Humean analysis of causality is correct - come to share in our adult delusions.”
(Kress, p.171)

And, in his final chapter, Kress turns explicitly to the broader intellectual climate, making a good case that we have over-extended the word “literacy” past the point of vacuity, pursued both formalism and its opposite into their respective dead ends, and - in “postmodernism” - frequently valued obfuscatory jargon ahead of any genuine form of understanding. Last, but certainly not least, he insists upon the continuity of writing (as sign-making) with the metaphoric springs of play...a crucial point which is all too easily forgotten in our classrooms. In all of these, he shows rare good sense, aligned to the sort of cross-disciplinary awareness most needed today, amidst the curse of over-specialization.


“A lack of attention to the formal aspects of literacy denies us essential knowledge. It denies us the possibility of knowing what precisely the maker of a sign meant at the moment of making the sign; it denies to us knowledge of her or his ‘interest’, which can reveal his or her self-perceived location in the social/cultural world. Equally significantly, it denies us the possibility seeing the process as one of the making of signs, rather than the using of signs...[for,] in my view, each use of the resources of literacy - whether in writing or in reading - results in the making of a new sign, even in the most ordinary of circumstances.... [So,] clearly an understanding of the formal aspects of literacy is crucial. At the same time...a focus on form inevitably invites a static view of form, or predisposes the theorist towards a reification of the medium and of the elements with which she or he is dealing.... These are serious problems.”
(Kress, pp.202-7)

“One fundamental characteristic of the postmodern theoretical turn is to show us fragmentation, complexities, multiplicities, parodies, and the impossibility of any real anchorage. To the extent that this is an apt description of contemporary social and cultural life, I would wish any theory of literacy, and the practices derived from it, to be fully attentive to these facts and, where necessary, remade. Where postmodernism is simply an ideological intensification and a mystification of existing states of affairs, as I believe it to be in many domains, I wish to resist such moves.”
(Kress, p.213)

“Learning of writing and reading - in school, or perhaps earlier, at home - is in no way [children’s] first encounter with the making of signs. It is however their first encounter with what will appear to them as arbitrarily constructed, unmotivated signs.... [Therefore,] a better way of thinking about this process is as the move from a period of nearly unconstrained production of metaphor, in all conceivable modes, into the much narrow range of modes which school validates. Above all, it is the move into the narrow and initially non-metaphoric mode of (verbal) literacy. There is a channelling towards culturally salient, and formal, modes of representation...[but] the starting point of theories of reading and writing, of literacy, must be the child’s disposition to the construction of metaphor. For children, that is what representation is. As it happens, and as always, they are correct in their assumptions and in their practices.”
(Kress, pp.219-21)

Gunther Kress’ Learning to Write (1994) is unlikely to ever graze upon the heights of the bestseller lists...even in the most learned of communities. And yet, it presents us with a cogently argued and very well-supported position which makes nonsense of so much of our so-called “linguistic turn” - based as that is (in all its dominant varieties) upon highly abstracted models with very little real relevance to the forms of language which surround us.

And, yet again, we can also see consilience developing - this time between psychology, linguistics, and key cultural historical approaches...as well as the rhetorical thinking so well embodied in the work of Richard Lanham. However, given Kress’ educational focus - as well as his strong  cross-disciplinary leanings - perhaps the most obvious linkage to be made would be with Kieran Egan’s developmental take upon educational theory, which (interestingly enough) lacks precisely the strong connection w/useful linguistic theory that Kress can supply. Moreover, as we will see in the following - and final - quotation, Kress insists upon connecting such with the many other forms of competence & understanding we have ignored, in our mistaken fixation upon the reified picture of language which is langue. All up, this is a rare book and - despite its resolutely academic address - one that deserves a much wider readership...


“Sound exists time; speech sounds are continuous, one sound merges into and affects the next. Language in the medium of sound is expressed through a fluid, continuous, temporal medium. Letters exist in space (e.g. on a page); letters are discontinuous, they are clearly segmented. Language in the medium of writing is expressed through a rigid, segmented, spatial medium. The constraints and possibilities of time on the one hand, and those of space on the other, attend language in either mode.... Just as one example, in writing it is possible to ‘browse’; in speech it is not...[but,] compared to language in the spoken form, the lettered medium strips away, cannot represent, speed, rhythm, intonation, voice quality, tone of voice (jocular, sarcastic, admonishing); it can only report them.... Compared to language in the spoken form, the lettered medium can and does make use of the compositional possibilities of a spatial medium, from layout to paragraphing, to the internal complex syntactic compositional arrangement of sentences.... The constraints, as much as the possibilities, of the medium of expression - letters - have an effect on the organization of the meaning system of language itself...nominalization, hierarchical sentence syntax, heavy pre-modification of nouns; all those forms which tend away from time and sequence, towards the spatial/static, and the compositional/hierarchical. [Furthermore,] one so far unstated and therefore unexamined corollary of attending to the specificities of literacy is that it is essential to attend to the specificities of other modes of representation, the visual, the aural, the musical, the gestural, [for] we cannot know what verbal literacy is unless we know not only how it differs from speech, but also how it differs from the visual, the gestural. The medium of expression together with its meaning system form a couple, whose possibilities of meaning are intimately bound up with the possibilities of the formal medium of expression. Attitude is probably more richly and variably expressed through the media of facial expression, or gesture or bodily posture, than through the written language.... This is why it matters absolutely what mode of representation is being used; heard, seen, or read; or analyzed. Not everything can be said in any medium...[and] meaning systems other than language have, so far, been so little studied, explored, theorized, that we have no commonly accepted names for them.”
(Kress, pp.209-12)



John Henry Calvinist