speech and writing linguistics

• Reference Manager
• Simple TEXT file

People also looked at

Original research article, the third dimension. on the dichotomy between speech and writing.

• Faculté des Lettres, University of Lausanne, Lausanne, Switzerland

This paper introduces a more complex and refined articulated view than the classic and simple dichotomy of linguistic production. According to the traditional doxa, what is linguistically articulated is either spoken or written. Forms of written language have previously been considered a secondary representation of spoken forms and, at least in the alphabetic system, the only properly linguistic form. I argue that there exists a third dimension of language, which is internal. This internal form is lexically, phonetically and grammatically articulated, without being spoken in a proper sense, but which can be seen as the pre-condition for both spoken and written production. In other words, linguistic production does not necessarily imply the presence of two interacting speakers (or writers/readers). Production can be seen as the simple effect of an internal activity, and can be described without reduction to spoken or written forms. A consideration of this third dimension in a systematic way could enrich and strengthen approaches to many types of texts and help to productively integrate the traditional schemes adopted in Sociolinguistics, Historical Linguistics, Philology, Literary Criticism, and Pragmatics.

Introduction

Speech in classical linguistic doctrine: saussure.

According to Saussure’s Cours de linguistique générale (from here on CLG , 27 Saussure, 1967 ), the act of parole is an individual one, but is realized as the minimum requirement of two “people who are speaking”:

“Pour trouver dans l’ensemble du langage la sphère qui correspond à la langue, il faut se placer devant l’acte individuel qui permet de reconstituer le circuit de la parole. Cet acte suppose au moins deux individus; c’est le minimum exigible pour que le circuit soit complet. Soient donc deux personnes, A et B, qui s’entretiennent”: Figure 1 .

FIGURE 1 . “Le circuit du langage” (from Saussure 1967 : CLG 60).

Thus, ideographic systems of writing directly represent the idea of words, and phonetic systems represent their sound ( CLG 47 ss.). Consequently (alphabetic) writing is the representation of the sounds of words, which is manifested in the act of a closed circuit shown above, which assumes two interlocutors. According to Saussure, there exists a connection from a concept to an acoustic image, then to phonation, and finally in inverse order, from a reassociation of the sound to an acoustic image, and then back to the concept Figure 2 .

FIGURE 2 . Phases of saussurean Circuit ( Saussure 1967 : CLG 60).

The written dimension is subordinated, both ontogenetically and phylogenetically, with respect to speech (the latter is identified with language tout court , with an almost imperceptible but crucial deviation). One reads in the CLG 45:

‘Langue et écriture sont deux systèmes de signes distincts; l’unique raison d’être du second est de représenter le premier; l’objet linguistique n’est pas défini par la combinaison du mot écrit et du mot parlé; ce dernier constitue à lui seul cet objet. Mais le mot écrit se mêle si intimement au mot parlé dont il est l’image, qu’il finit par usurper le rôle principal; on en vient à donner autant et plus d’importance à la représentation du signe vocal qu’à ce signe lui-même. C’est comme si l’on croyait que, pour connaître quelqu’un, il vaut mieux regarder sa photographie que son visage ( CLG 45).’

Typical of pre-nineteenth century linguistics was the centrality of writing and written language. Twentieth-century linguistics, then, pushed writing to one side, focusing on the only other perceived dimension, speech. Bloomfield’s famous statement (1935, p. 21) “Writing is not language” became necessarily integrated, in the American structuralist’s perspective, with the notion that the spoken dimension is the only one which duly qualifies as language .

From CLG 45 one can read an entire history of twentieth-century linguistics, which appears to have always taken for granted the dependence of written language on spoken language. This perspective is succinctly highlighted by Martinet, (1972) , p. 70): “a graphic code exists, writing, but apart from this there is no other code: there is language”, obviously referring to speech. There exists almost no twentieth-century treatize which does not define spoken and written language in terms of a dichotomy, and as being the primary (originally, only) and secondary (derived from the first) dimensions of linguistic activity respectively. And there is no work, even among the most recent and attentive studies to questions of the relationship between writing and speech, which does not tend to consider speech simply as the motor of innovation of writing, excluding interference or the role of any other dimension.

Ultimately, according to the model hypothesized by Saussure, spoken language is crucially super-individual. It presupposes at least two individuals, as discussed above. Alphabetic written language is simply a secondary (and often distorted) representation of spoken language, which constitutes the only object of linguistics properly understood (that is, the linguistics of langue , as per the explanation in CLG 38–39).

Twentieth-Century Criticism of the Dichotomy Between Speech and Writing

The dichotomy between speech and writing is discussed at several points during the 20th century. Strictly speaking, this dichotomy is not one of the greatest Saussurian dichotomies, given that Saussure does not theorize it with the same articulation with which he outlines other oppositions, such as those between Langue and Parole, or even Diachrony and Synchrony etc. One reason may be due to the fact that, from his point of view, the written dimension is simply external to the field of linguistics. Already from a structuralist perspective à la Hjelmslev (1966, pp. 131–32), in fact, we see how written and spoken language are not derived from each other, but are simply two manifestations of the same form. In this account, the priority for speech over writing had already been questioned—not from a historical point of view, but from an axiological and epistemological one.

Among the most important discussions, there has been some attempt to refine the sharpness of the boundary between the two fields, highlighting the elements of continuity and, in part, intersection. This is the case of the Koch-Österreicher (1990) model: to the simple distinction between written language vs. spoken language, the two German Romanists oppose a model based on the concepts of distance vs. closeness. These concepts allow, on the one hand, a further realization of the sociolinguistic aspect of situations devoid of writing, and on the other hand, allow us to frame those phenomena which are clearly mixed or hybrid.

The Koch-Österreicher model proposes a scale of distance (and of the quantity of interlocutors included by the single linguistic act) whose minimum value is in fact 1. The conditions of communication are identified, in the first instance, in the Grad der Öffentlichkeit («für den die Zahl der Rezipienten—vom Zweiengespräch bis zum hin zur Massenkommunikation», Koch-Österreicher 2011, p. 7, Italics mine). In short, as in the Saussurian model, there does not appear to be an inferior degree with respect to communication when two are present.

Linguistics in the late 20th century elaborated the concept of diamesic variation (a term invented by Mioni 1983 , extending a series of analogous categories from Coseriu). But it struggled to demonstrate that there exist various “intermediate” positions between these two poles. Apparently, the poles are not united (as instead occurs for similar polarities, such as the classic dimensions of sociolinguistic variation).

A further contribution to overcoming the exclusive and rigid dichotomy between writing and speech was provided by the twentieth-century development of studies on sign language (SL). It is thanks to this line of research and the continued appreciation of SL as an alternative channel to spoken and written language, that traditional expressions such as “spoken or written language” are often substituted with other ones. In recent studies, a trinomial “spoken, signed or written language” (for example, as recently as in Haspelmath 2020 , p. 2) has entered the literature. In short, one finds an all-encompassing category of verbal language in addition to the traditional dichotomy of spoken/written language. This category synthesizes, rather than supersedes, the old contraposition (notwithstanding the distinct nature of signed languages, which can be acquired spontaneously, with respect to writing, and which are the fruit of cultural transmission and learning).

In sum, twentieth-century linguistics approaches the polarities of written vs. spoken language both in a theoretical perspective as well as in a specifically sociolinguistic perspective. Linguistics aimed to overcome this conception as an exclusive dichotomy, placing greater emphasis on those elements of continuity which overlap. This approach was favored also by the emergence of new methodologies of communication. A further contribution to superseding the written/spoken duality was provided by research on sign language: dealing, as it does, with phenomena that cannot be reduced to either category, nor to either one of the polarities.

The Internal Text of G. R. Cardona

Among the few contributions which properly highlighted the linguistic question posed by an internal text, and taking inspiration from both literary and non-literary texts, is the work of Giorgio Raimondo Cardona, (1986 , reprinted Cardona, 1990 ). Published 2 years before his unexpected and premature death, the article deals with “mental text” as an indispensable premise for the production of any oral, but especially written, text.

Let us consider in particular one of the crucial passages of Cardona’s essay: “There is no analytic thread (whatever its itinerary may be) that can be exempt from choosing internal discourse as a point of departure: apart from some cases of automatic writing or trance or similar, no external communicative activity can be disregarded from endophasic, mental and communicative discourse”.

Cardona, (1986) begins from an examination of the literary manifestations of internal speech, bringing to light suggestions from the field of semiotics (and particularly from Lotman et al., 1975 ). He focuses on criticism from genetics and on twentieth-century variationist linguistics, before moving to what he considers a particular type of text, understood as a preparatory and evolving phase that precedes the development in written form, but also its spoken realization. In this way, “the various genres, written and spoken, open up into a natural typology, widening to become waves from the nucleus of internal discourse”.

Another fundamental passage from Cardona consists in recognizing internal discourse (or “interior text”) as the essential absence of a pragmatic dimension, beyond an extreme simplification of syntax (“in one’s thought for oneself, the combination can be reduced to its minimum, the mental nuclei find their minimal linguistic expression. It can, at times, be substituted or integrated by images, as per the manuscripts of Leonardo da Vinci”). Cardona’s examples usefully extend to textual and typological instances that are quite varied.

Cardona’s gifted intuition has been recognized by Italian studies of general linguistics, by history of the Italian language, and literary criticism, which occasionally quote him (among the most important studies, see D’Achille 1990 : 18, who dedicates a note to him). But it has never been explored in full and, in fact, it has not led to any substantial new analysis in the general study of written and spoken language. Significant, for example, is the absence of any reference to him in the best work of German Romance studies in the new century, from Kabatek (2000) to the second edition of Koch-Österreicher (2011) . Furthermore, Cardona’s work does not appear to have been recognized even by contemporary French linguistics, which has, on several occasions, returned to the notions of langage and parole intérieur(e). This includes within the traditional studies of psychology, which we will take up below (exemplary in this regard is Bergounioux 2001 ). In terms of Italian linguistics, which has always been attentive to the social dimension of language and its recent evolution, historians of the Italian language have concentrated mainly on the opposition between spoken and written speech (see, for example, the studies following the work of Giovanni Nencioni, later published in Nencioni 1983 ). Up until now, Cardona’s work has mainly influenced the realm of literature (for example, Bologna 1993 ).

The Perspective From Generative Linguistics

Linguistics in the past few decades has opened up a debate with particular vigor, especially in the field of studies on the origin of language and its biological foundations, which can be summarily characterized by the following two extremes: 1) language is studied primarily as an instrument of communication; 2) language is studied primarily as a form of organization of thought.

Generative grammar resolutely derives from 2) within a theory of externalization. This theory identifies the human specificity of language in its internal, computational (syntactic) capacity (that is, in what Chomsky calls internal language, I-language) and not in the interaction between it and the materiality of phonation. As for the executive function of neurons, the human species shares this aspect with various other animals (hence the computational-syntactic capacity is exclusive of homo, cf. for example Berwick, 2013 ). Generally speaking, only syntactic characteristics are assigned to I-language, dealing as it does with a computational system, that is, with the product of a mental apparatus.

In a partially complementary position, a recent string of neurolinguistic studies (for example, the various works by A. Moro and others, cf. Magrassi et al., 2015a ) has made it possible to observe the cerebral traces of the mental representation of words with the tools from clinical observation. These observations occur not just during the listening phase, but also in the phase of production. In particular, they are visible in those areas of the encephalon that are crucially non acoustic, such as the Broca area. This has highlighted the many affinities between spoken language and “thought language”: the latter showing a great number of elements in common with the former, and thus comprising something similar to what Saussure had already called the acoustic image of words. In short, to summarize with an efficient phrase from Moro, (2016) , p. 89: “when we think without speaking, we are putting the sounds of words in our thoughts”.

One consequence of the theories and of the hypotheses (even though partially divergent) of what we have just said, is that recent linguistics has made it possible to ascertain a certain finding of linguistic dimension preceding phonation, but still within the domain of linguistics. This is due to the fact that we are dealing not only with syntactic structure, which must be considered the specific nucleus of the very faculty of language, but also of a phonetic and phonic consistency at the level of the neural networks. It is, therefore, a recognition in terms of the language(s) involved. In other words, we do not only think linguistically but (at least in certain situations), rather in a very well defined language .

Therefore, not only does language have a foremost interior dimension, if it is understood as the disposition of a computational system with a mainly syntactic nature. It also has a further dimension, still internal, but to which we can add the application of universal syntactic parameters as well as characteristics that are already fully recognizable as single languages. There exists, that is, a form of thought which is already proprerly articulated (and is even formed with features of a single language). At the same time, it is independent from an external, phonic expression in the same language, with which it also maintains strong relations even at the level of activation of neural networks linked to hearing.

The Perspective From Textual Criticism and Philology

The existence of forms of linguistic production that are independent both from acts of phonation, and from writing, has always been known in an intuitive sense. Nevertheless, the received wisdom has tendentially merged or even confused the mental articulation of language with the dimensions of speech or writing. This appears to be the case with metaphors of daily language such as “I said to myself” (in order to introduce the content of a thought that is not truly “said”, but simply “thought”), or “I made a mental note”. Therefore, mental content exists that is linguistically articulated but which is neither “said” nor “written” in the true sense of these two words.

Even literary production has always considered the purely internal dimension of language, and not in a written or spoken sense. Literary works have given conventional representations which, once again, are mainly anchored in the traditional forms of dialogic speech: the form of an (interior) monologue assumes, in an earlier literary tradition, elements such as allocution to oneself (such as those of epic or tragic heroes, as well as lyricists of Antiquity). These elements represent the endophasic dimension as a variation of speech in which two interlocutors coincide.

More recent forms of literary representation (for example, the twentieth-century stream of consciousness) have allowed an attempt to give an autonomous and more “realistic” representation of such phenomena to emerge. In recent years, French stylistique has deepened the literary reflexes of the late nineteenth-century psychological debate (especially in France, as detailed below) on langage intérieur (for example, Rabatel 2001 , Martin-Achard 2016, Dujardin 1931 , Pettenati 1961 ). The in-depth analyses of literary criticism are numerous in these fields (for example, Philippe 2001 ; see above for G. R. Cardona’s particular linguistic view on variationist linguistics).

In fact, the (at least) partially autonomous nature of the articulation of internal language appears to have slipped away attention from its spoken form. This does not mean they are completely separated from it. Little attention has been given to the fact that the same act of writing (autonomous or as a form of copying) assumes a formalized pre-elaboration of content which is not spoken at all, but only thought.

In recent times, before the neurolinguistic studies discussed above, even a particular phenomenon such as transference— via copying—of a written text to another written text has been studied within philology. Indeed, philology has considered phenomena such as the so-called internal dictation in a profound way during the course of the 20th century (see the fundamental studies by Alphonse Dain (1975) ; on “internal pronunciation”, and cf. also Avalle’s considerations 1972, p. 34). Philology has identified a great number of indices which refer us back to a form of “listening” and internal “repetition” (in an acoustic sense) of a graphic sequence that is looked at during the first act of copying, and then transcribed in the second. Most copying errors that are ascribable to defects of internal dictation can be traced, in fact, to the acoustic nature of such repetitions. These errors, nevertheless, do not assume any sound if only that “of thought”, to return to Moro’s expression.

The Perspective From Psychology

The category of internal language was investigated in the fields of psychology and medicine before linguistics. The research conducted by Victor Egger (1881) , Egger (1904) and by Georges Saint-Paul (1892) , Saint-Paul (1904) , Saint-Paul (1912) during the last twenty years of the 19th century, has a seminal value. Partly adopting contrasting perspectives, they proposed establishing a typological classification of the forms of endophasy, bringing attention to the faculty of hearing, as well as visual and verbal-kinetic aspects of the internal representation of language. Until then, these aspects were not able to be investigated simply through introspection (Egger) or interrogation of witnesses (Saint-Paul). The latter originally used a questionnaire which was also distributed among writers; for a historiographical overview of this debate, see Carroy 2001 .

As is general for other aspects of linguistics, the approach that is based on the study of child language acquisition has allowed us to untangle that which appears difficult to ascertain in adults. According to Lev Vygotsky (1966) , whose theory on the formation of internal language is widely accepted, language in the child has a function for social interaction with people in immediate surrounding. Then an egocentric phase from which socialized and internal language derive.

In this particular area, Vygotsky’s model is accepted in substance by Jean Piaget, while reinterpreting the Vygotskian theory. Despite some cases of divergence on specific points, even the theoretician of genetic psychology agrees with the hypothesis that egocentric child language is the point of departure for the development of internal language. This phase is found during a successive stage of development, and is parallel to the formation of “socialized” language (it does not follow it, therefore, and is not derived from it either).

To the general category of internal language can be traced, in the adult, both endophasy (which does not assume any phonation), as well as solitary speech, which represents a sort of medial point between the proper dimension of internal language and the typical dimension of speech in the presence of an interlocutor.

Despite the debates outlined above in the field of psychology, a conspicuous part of general linguistics has continued (more or less) to explicitly reduce internal language to a simplified form of dialogue, in which two interlocutors coincide, through a sort of duplication of the subject into two interlocutors. In fact, this is the point of view, for example, that Benveniste, (1974) , p. 85 adopts in considering monologues: “le monologue procède bien de l’énonciation. Il doit être posé, malgré l’apparence, comme une variété du dialogue, structure fondamentale”. The example is valid also in showing a much broader tendency as well.

Writing, Speech, Thought

In reality, it is obvious that most linguistic production however it is understood occurs outside the domains of speech and writing. Most content that is articulated in a linguistic way (and, as we have said, this includes also mental content, in every sense) happens in thought, and precedes—literally—any form of external expression, spoken or written.

The way in which language is articulated internally is still largely unattainable. This explains the reason why its perception has always turned out to be fleeting, and its nature confused with other forms. If this is the case, the same relationship between spoken language and written language has been read in a completely different way from other graphical cultures (for example, Chinese, on which see the recent paper by Banfi 2020 ). This relationship has been consistently characterized by a tradition that adopts graphemes of a phonetic nature, and in a modern way.

The fact that “thought” language is attainable only in a difficult way, and describable only in specific forms, does not mean that it does not exist, however. The recent findings from neurolinguistics (which have created the possibility of tracing the recognition of syntactic structures independent from sound in the brain, as Moro et al., (2020) have recently done) open up interesting perspectives on the concrete attainability of the thought dimension of language. But even other elements may be involved, in the same sense.

On the other hand, even the spoken dimension of language (obviously much more relevant than written language) has long been neglected, since it is more difficult to obtain with respect to written language. Today, speech can be observed in various forms–that is, one can not only transcribe, but also record. This means that it has been considered as an autonomous subject.

Describing the study of the thought dimension, even in linguistic studies, could have further consequences for the way in which the two tangible dimensions of writing and speech are evaluated. We know that these dimensions influence each other. Koch and Österreicher have produced the most refined model, perhaps, to describe such reciprocal influence. But we do not know exactly what the relationship between them and the third dimension is.

The representation of the relationship that has traditionally been conceived between speech and writing can therefore be summarized in a simple dependency of one on the other:

SPEECH(language proper).↓Writing.(conventional representation of speech, «non language»)

The model proposed by Koch-Österreicher energizes and complicates such a representation, while maintaining an eminently communicative vision of language, shifting the focus toward the notions of Distance and Conception: Figure 3 .

FIGURE 3 . Spoken vs Written Language according to Koch-Österreicher 1990 : 13.

Cardona (1986) adopts an even broader and more articulated perspective, producing a model that is formally similar to those that were being elaborated contemporaneously in various subdisciplines of sociolinguistics (e.x., the well-known Berruto, 1987 model for contemporary Italian): Figure 4 .

FIGURE 4 . Spoken vs Written Language (and Internal Language) according to Cardona (1986) .

The direction which Cardona had already invited us to consider, and which the combined perspectives of twentieth-century psychology and recent neurolinguistics appear to endorse, is that of an even more decisive integration of the internal dimension in the study of language and languages. The consideration of thought language appears to be inevitably presupposed to the study of every manifestation—spoken and written—of language itself. One can attempt to supersede, in this way, the traditional, hierarchical vision which subordinates speech to writing on both of the traditionally identified dimensions. Both appear to be subjected, and equally so, to the overriding internal elaboration of language.

In this sense, the persistent idea loses some force that speech should take on a priority role in both the description and the realization of language. Speech is, certainly, the most direct and immediate projection of thought, but perhaps not the necessary cause of every manifestation of writing. Rather, in many cases it continues from thought in a much more plausible way. Naturally, this does not prevent the idea that the conception (in the sense intended by Koch-Österreicher) of text can bring the dimensions of speech and writing into communication. In the views which we have summarized here, they do not necessarily seem to be in a relationship of direct derivation: Figure 5 .

FIGURE 5 . Thought, Speech, Writing according our hypothesis.

In this model, the different length of the sides of the triangle refer to the diverse nature of spoken forms, as is natural, with respect to written forms, which are cultural and obviously negotiated. A possible representation of what is signed in this model would lead to a further derivation from thought which, in turn, would be independent from speech.

Possible Prospects: Outside Literature

Among the consequences of a possible autonomous recognition of the thought dimension of language, distinct from both speech and writing, is the question of overcoming an automatic act which appears to be reasonably widespread. This act derives from the consideration of written language as a simple reflection of speech. If what Saussure had already observed concerning alphabetic writing is irrefutable (that is, that writing reproduces more or less efficiently the phonic substance of words), then it is also true that a certain tendency can be observed to attribute the least characteristic or marginal facts of written language to a pure and simple influence of speech. In the elaboration of writing, thought generally appears much more decisive than speech proper.

In reality, there are various forms of written production that are difficult to interpret as representations of corresponding forms of speech. At the most simple level, what is obvious in forms of elementary text (the oldest attested forms in the development of writing) such as lists, notes, or annotations written down from memory, the writer does not address others but rather him or herself. Nor does the author intend to be comprehended by people other than themselves. It will be useful to recall that among the earliest manifestations of writing throughout history, we find functional texts that are not intended for interpersonal communication, nor for the reproduction of spoken discourse, but rather computational ones. In other words, numbers are born well before letters and “the code of abstract ideas, in particular the numerical code, seems to have performed an essential role from the first stages in the appearance of writing, and perhaps in the very idea that concepts can be written down” ( Deahaene 2009 , p. 211).

In general, a large part of so-called “semi learned” texts, which have been the object of linguistic enquiry for just a short time, present a linguistic phenomenology that is perhaps inappropriately described as being influenced by speech. A much more persuasive explanation of its various phenomena is provided by referring to the dimensions of thought, rather than to speech.

Cardona (1986 , p. 80) has also investigated this aspect of language. With respect to the category of ‘semi learned’ persons, he alludes to the modality of “writing down in real time one’s own mental discourse which is first and foremost—due to a lack of other models—an oral discourse”. But the priority of oral discourse appears dominant here too, when it appears necessary to shift toward a description of the syntax of thought in an analogous way that, for the syntax of speech, has allowed us to re-read and re-interpret such phenomena of written production coherently (as well as programmatic, in this sense, see Sabatini 1990 ).

Therefore, it can be useful to reconsider in a systematic way the elements which in non-literary writing (and particularly in less attentive writing) have traditionally been considered as reflexes of spoken language. One may ask whether these elements should not be removed from that dimension, and restored to the proper category of internal language. To quote one of the clearest and most recent formulations, it is a common opinion that “semi learned writing is characterized by an integral and large adoption of spoken structures” ( Testa 2014 , p. 107). But semi learned writing also includes the modality that Trifone 1986 has aptly described as being “writing for oneself”. Whether this type of writing simply integrates elements and styles of spoken language is a partially equivocal notion. It is created by a lack of features, made up of traits that are external to spoken language, and includes elements both of speech proper and elements of thought. Diaries, notes, jottings made out of necessity or from memory: those who write for themselves (and more so if the writer is semi learned) do not necessarily rely on speech, but more likely draw on the most immediate form of their linguistic production: thought.

Possible Prospects: Literary Production

Some elements of thought language have been highlighted by criticism and literary theory. But in terms of linguistic studies applied to literary texts, there seems to remain a certain reluctance to consider the relationship between internal language and literary language.

We have often borne witness to an appreciation of the literary reproduction of speech. In other words, the mimetic capacity of some literary production (especially in prose) reproduces phenomena in written form that are (or would be) unique to orality. This is one line of research that has been very productive, and which has the merit of clearly distinguishing between that which pertains to the written dimension (studied longer and in a deeper way) from that which does not pertain to it. In a certain sense, it is as if the term “speech” has long indicated simply “that which is not written”, or whatever is different from writing.

Among literary texts that best lend themselves to an indirect investigation of the typical characteristics of thought language, we find also poetic texts (especially lyrical ones), in which the subject, at the center of the discourse, does not seem to have any interlocutor. These texts can be placed alongside prose, discussed above on the flow of conscious and internal monologue.

Economy of syntax, omissions of references to context, advances of the text free from association of ideas without explanations, and a centering of the ego: these are just some of the elements which distinguish a part of poetic production—particularly modern poetry—from more traditional forms of poetic discourse, founded above all on the adoption of canonical verse and metrics. A conspicuous part of modern poetry seems to distinguish itself from prose above all for its privileged link, even if implicit, with the internal dimension of language, that is with thought language.

In this way, some stylistic traits unique to poetry—especially recent poetry—can be explained in an even more persuasive way if one attempts to describe them as outcomes just in terms of thought language. These forms have been typically characterized as an implausible reproduction of speech. This is a parallel, but distinct, step with respect to what we have said above in terms of non-literary writing. In both cases, it is a question of overcoming the almost seamless, and unwarranted, process of assigning phenomena that occur in certain forms of written language only in a marginal and peculiar way to an implausible flow of speech.

In conclusion, the intersection between literary writing and thought language deserves to be explored more attentively, with tools appropriate to linguistics. The noteworthy study of tracing reproducible elements, more or less consciously, of speech in literary texts could also be applied in identifying elements of thought language in literary writing proper. In modern poetry, the ongoing relaxation of the canonical, formal requirements seem to be compensated by an ever stronger relationship between poetry and thought language, whose syntactic, textual, and pragmatic points deserve further definition and articulation.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author Contributions

The author confirms being the sole contributor of this work and has approved it for publication.

Conflict of Interest

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Avalle, D.-S. (1972). Principi di critica testuale . Padova: Antenore .

Banfi, E. (2020). “ Antinomia ‘parlato’ vs. ‘scritto’ Nel Pensiero Linguistico Cinese ,” in L’antinomia scritto/parlato , a cura di Franca Orletti e Federico Albano Leoni, Città di Castello, ( Emil , 69–87.

Google Scholar

Benveniste, É. (1974). Problèmes de linguistique générale . Paris,: Gallimard .

Bergounioux, G., (2001). (dir.), La parole intérieure , volume thématique de “Langue française” , 132.

Berruto, G. (1987). Sociolinguistica Dell’italiano Contemporaneo . Roma: La Nuova Italia .

Berwick, R. C., Friederici, A. D., Chomsky, N., and Bolhuis, J. J. (2013). Evolution, Brain, and the Nature of Language. Trends Cogn. Sci. 17, 89–98. doi:10.1016/j.tics.2012.12.002

PubMed Abstract | CrossRef Full Text | Google Scholar

Bloomfield, L. (1935). Language . London: Allen & Unwin .

Bologna, C. (1993). Tradizione e fortuna dei classici italiani: Dall'Arcadia al Novecento . Torino: Einaudi .

Cardona, G. R. (1983). “Culture dell'oralità e culture della scrittura,” in (a cura di), Letteratura italiana , II, Produzione e consumo . Editor A. Asor Rosa (Torino: Einaudi ), 25–101.

Cardona, G. R. (1990). I linguaggi del sapere , a cura di C . Bologna, Roma-Bari: Laterza .

Cardona, G. R. (1986). Testo Interiore, Testo Orale, Testo Scritto. «Belfagor» 41/1, 1–12.

Carroy, J. (2001). Bergounioux 2001 , 132, 48–56. doi:10.3406/lfr.2001.6314Le langage intérieur comme miroir du cerveau : une enquête, ses enjeux et ses limites “Langue française”

CrossRef Full Text | Google Scholar

D’Achille, P. (1990). Sintassi del parlato e tradizione scritta della lingua italiana . Roma, Bonacci: Analisi di testi dalle origini al secolo XVIII .

Dain, A. (1975). Les Manuscrits . Paris, Les Belles-Lettres .

Deahaene, S. (2009). I Neuroni Della Lettura , Trad. it . Milano: Raffaello Cortina .

Dujardin, E. (1931). Le Monologue Intérieur . Paris: Messein .

Egger, V. (1881). La parole intérieure. Essai de psychologie descriptive . Paris: Berger-Levrault .

Egger, V. (1904). La parole intérieure. Essai de psychologie descriptive . Paris: Alcan .

Haspelmath, M. (2020). Human Linguisticality and the Building Blocks of Language. Front. Psychol. 10 (January 2020), 3056. doi:10.3389/fpsyg.2019.03056)

Hjelmslev, L. (1966). Prolégomènes à une théorie du langage , trad. fr. Paris: Editions de Minuit , 1968–1971.

Kabatek, J. (2000). L’oral et l’écrit – quelques aspects théoriques d’un «nouveau» paradigme dans le canon de la linguistique romane, in W. Dahmen, G. Holtus, J. Kramer, M. Metzeltin, W. Schweickard, and O. Winkelmann (Hrsg.), Kanonbildung in der Romanistik und in den Nachbardisziplinen . Romanistisches Kolloquium XIV , Tübingen, Narr , pp. 305–320.

Koch, P., and Österreicher, W. (1990). “Gesprochene Sprache in der Romania”, in Französisch, Italienisch, Spanisch, Tübingen: Niemeye. doi:10.1515/9783111372914

Koch, P., and Österreicher, W. (2011). “Gesprochene Sprache in der Romania,” in Französisch, Italienisch, Spanisch , 2 . Ausgabe (Berlin/New York: De Gruyter ). doi:10.1515/9783110252620

Lotman, J., Uspenskij, M., and Boris, A. (1975). Editor R. M. Facciani / M. Marzaduri, Tipologia della cultura , Milano: Bompiani.

Magrassi, L., Aromataris, G., Cabrini, A., Annovazzi-Lodi, V., and Moro, A. (2015a). Sound Representation in Higher Language Areas during Language Generation. Proc. Natl. Acad. Sci. USA 112 (6), 1868–1873. doi:10.1073/pnas.1418162112

Magrassi, L., Cabrini, A., Aromataris, G., Moro, A., and Annovazzi Lodi, V. (2015b). “Tracking of the Speech Envelope by Neural Activity during Speech Production Is Not Limited to Broca’s Area in the Dominant Frontal Lobe,” in Articolo presentato durante la 37th Annual International Conference of the ieee Engineering (Milano: Medicine and Biology Society ).

Martinet, A. (1972). “Lingua parlata e codice scritto,” in Linguistica e pedagogia , trad. it . Editor J. Martinet (Milano: Angeli ), 69–77.

Mioni, A. M. (1983). Italiano tendenziale: osservazioni su alcuni aspetti della standardizzazione, in Scritti linguistici in onore di Giovan Battista Pellegrini , Pisa: Pacini , 1983, pp. 495–517.

Moro, A. (2016). Le Lingue Impossibili . Milano: Raffello Cortina. doi:10.7551/mitpress/9780262034890.001.0001

CrossRef Full Text

Moro, A., Micera, S., Artoni, F., d'Orio, P., Catricalà, E., Conca, F., et al. (2020). High Gamma Response Tracks Different Syntactic Structures in Homophonous Phrases. «Scientific reports» . https://www.nature.com/articles/s41598-020-64375-9 .

Nencioni, G. (1983). Di Scritto e di Parlato. Discorsi linguistici , Bologna: Zanichelli.

Pettenati, G. (1961). Su alcuni riflessi del discorso interiore nel discorso scritto. «Annali dell’Istituto Universitario Orientale – Sezione Linguistica» 3, 237–246.

Philippe, G. (2001). Le paradoxe énonciatif endophasique et ses premières solutions fictionnelles, in Bergouioux 2001 ( “Langue française” , 132), pp. 96–105. doi:10.3406/lfr.2001.6317

Rabatel, A. (2001). Les représentations de la parole intérieure [Monologue intérieur, discours direct et indirect libres, point de vue], lfr, in Bergounioux 2001 ( “Langue française” , 132), pp. 72–95. doi:10.3406/lfr.2001.6316

Sabatini, F. (1990). Una Lingua Ritrovata: L’italiano Parlato [1990], Ora in Id., L’italiano nel mondo moderno. Saggi scelti dal 1968 al 2009, ed. by V. Coletti, R. Coluccia, P. D’Achille, N. De Blasi, D. Proietti, Napoliet al. 2011, pp. 89–108.

Saint-Paul, G. (1892). Essais sur le langage intérieur . Lyon: Storck .

Saint-Paul, G. (1912). L’art de parler en publique . Paris: L’aphasie et le langage mentalDoin .

Saint-Paul, G. (1904). Le langages intérieur et les paraphasies . Paris: Alcan .

Saussure, F. (1967). Cours de linguistique générale . Paris: Payot .

Testa, Enrico. (2014). L’italiano nascosto. Una storia linguistica e culturale . Torino: Einaudi .

Trifone, P. (1986). “Rec. a E. Mattesini, Il ‘Diario’ in volgare quattrocentesco di Antonio Lotieri di Pisano notaio in Nepi (1985),” in «Studi linguistici italiani» XIIRinascimento dal basso. Il nuovo spazio del volgare tra Quattro e Cinquecento ( Roma: Bulzoni ), 133–142.

Vygotsky, L. S. (1966). Pensiero e linguaggio (con un’appendice di Jean Piaget) . Firenze: Giunti .

Keywords: language, spoken (and written language), written, psycholinguistic, linguistic variation

Citation: Tomasin L (2021) The Third Dimension. On the Dichotomy Between Speech and Writing. Front. Commun. 6:695917. doi: 10.3389/fcomm.2021.695917

Received: 15 April 2021; Accepted: 06 May 2021; Published: 07 June 2021.

Reviewed by:

Copyright © 2021 Tomasin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Lorenzo Tomasin, [email protected]

Linguistics: Between Orality and Writing

• First Online: 22 August 2022

Cite this chapter

• Harvey J. Graff 2  

255 Accesses

In the beginning, there was the Word. The word was spoken. Our knowledge of this comes through writing. It also comes through centuries of translation and conflicting interpretations: a set of relationships that plagues understanding. We have a long legacy of formulaic divides surrounding “ from oral to written or literate” that assume a historical, evolutionary trajectory. The linguistic bases of literacy studies swing from presumption of antecedent to subsequent.

The basic study of language divides over the primacy and determinative influence of either the oral or the written. Linguistics’ roots in religion and foundations in philology are not appreciated.. This is part of their neglect—or the presumption—of history, and of their acceptance of the primacy of a foundational shift from oral to written.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Available as EPUB and PDF
• Read on any device
• Instant download
• Own it forever
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

I first used this rhetorical formulation in Legacies of Literacy (1987). The similarity between this formulation and the opening of the first chapter in Jack Goody’s Logic of Writing ( 1986 ) is entirely coincidental. Yet the differences between our interpretations of this logocentric view are consequential. While I immediately underscore the principal issues and sources of confusion in understanding literacy, Goody turns a complex historical transformation into a formula.

Brockmeier’s version of his “episteme” is deeply ahistorical; none of these formulations pays attention to context. Claims of its novelty are self-serving. Brockmeier and Olson’s use of evidence is flawed. They confuse and conflate social, psychological, and intellectual issues; the general and specific; and literacy and writing. For one critique of Olson and his presumptions, see Halverson, “Olson on literacy” ( 1991 ); Halverson, “Goody and the implosion of the literacy thesis” ( 1992 ).

Brockmeier and Olson’s “Literacy episteme” ( 2009 , 9), declares the existence of a “field” after 1960, but that is not the same as an episteme.

For an interesting perspective, included in one of the testaments to the literacy episteme that, looking to the “future of writing” rather than the past and present, actually argues somewhat contradictorily against part of the accepted narrative, see Harris, “Literacy and the future of writing” ( 2002 ). See also Cook-Gumperz and Gumperz, “From oral to written culture” ( 1981 ), Cook-Gumperz, Social construction , 1986 .

In “Are there linguistic consequences of literacy?” Biber argues: “For example, researchers such as O’Donnell (1974), Olson (1977), and Chafe (1982), argued that written language generally differs from speech in being more structurally complex, elaborated, and/or explicit” (Biber 2009 , 75).

See the classic work of Basil Bernstein, Dell Hymes, Erving Goffman, William Labov, and their students. For introductions, see Bauman and Briggs, “Poetics and performance” ( 1990 ); Goffman, Forms of Talk ( 1981 ); Hymes, “Introduction: Toward ethnographies of communication” ( 1964b ); Koerner, “Toward a history of modern sociolinguistics” ( 1991 ); Labov, Sociolinguistic patterns ( 1972 ); Shuy, “Brief history of American sociolinguistics,” 1990 ; Szwed, “Ethnography of literacy” ( 1981 ).

Finnegan quotes then Director-General of UNESCO, Réné Maheu, speaking of the “apparent association between non-literacy and illiteracy” and asserting “one apparent consequence of nonliteracy: lack of literature” ( 1973 , 113).

Among her major targets are Goody, Ong, Havelock, and McLuhan.

Yugoslavia was the name of the country when Lord published The Singer of Tales in 1960.

See for example, Lord’s chapters on “Writing and Oral Tradition” and “Homer” in Singers of Tales (1960). Compare on one hand with Havelock’s work and on the other hand with Finnegan, Literacy and Orality (1988). See also Lord, Singer of Tales , 2000 ; Parry, Making of Homeric Verse ( 1971 ).

Havelock and Marshall McLuhan were colleagues at the University of Toronto. Among the many influential works on the alphabetization of the brain that acknowledge a debt to Havelock is Maryanne Wolf’s Proust and the Squid ( 2007 ). Havelock’s work, which is pervaded by such slippages, is often cited on the “great transmission” or the revolutionary remaking of the human brain.

Compare with Lord, Singers of Tales (2000, 130) and the extended example of Yugoslav oral poets. Havelock, Literate revolution in Greece ( 1986 , 167), refers to “erosion of orality.”

Finnegan writes: “When we speak of both ‘orality’ and ‘literacy’ one or more of three main aspects may be involved: composition, performance, and transmission over time. These three do not always coincide. Thus it is possible for a work to be oral in performance but not in composition or transmission, or to have a written origin but non-written performance or transmission. These various combinations constitute a background to considering different patterns of transmissions…. The differing patterns do not coincide neatly with the distinction between oral and written traditions” (1988, 171–172). See Heath, Ways with words (1983); Schieffelin and Gilmore, Acquisition of literacy ( 1986 ); Tannen, Spoken and written language ( 1982 ). For examples of inattention to context and oral-literate relationships, see Canagarajah, Translingual practice (2013); Blommaert, Grassroots literacy ( 2008 ).

On letter writing, see popular South American films; Besnier, Literacy, emotion and authority , 1995 ; Cancian, Families, lovers, and their letters ( 2010 ); Henkin, Postal age, 2006 ; Kalman, Writing on the plaza ( 1990 ); Lyons, Readers and society , 2001 ; Lyons, Reading culture and writing practices ( 2008 ); Lyons, Writing culture of ordinary people ( 2013 ); Romani, Postal culture ( 2013 ); Scribner and Cole, Psychology of literacy ( 1981 ); Vincent, Literacy and popular culture ( 1989 ); Vincent, Rise of mass literacy ( 2000 ).

Contrast the Maori’s experience with that in Fiji described by Clammer, Literacy and social change ( 1976 ). According to Tagupa, “Education, change, and assimilation” ( 1981 ), missionaries and officials in Hawai‘i presumed that an alphabetic translation and print led directly to mass literacy and expected that cultural and social changes would necessarily follow.

For Central and South America, Salomon, “How an Andean ‘writing without words’ works” ( 2001 ), Hanks, Converting words ( 2010 ), and Rappaport, Politics of memory ( 1990 ) form excellent case study material. See also Seed, “‘Failing to marvel’” ( 1991 ). For great divide views, see Mignolo, Darker side of the Renaissance ( 1995 ); Boone and Mignolo, Writing without words ( 1994 ). For North America, recent scholarship on native peoples and their encounters with colonists informs the same fundamental questions and follows the same trajectory. Less sophisticated and less influenced by both linguistics and anthropology but now developing rapidly, Native American literacy studies has also been less influenced by scholarship in literacy studies. Regardless, it is ripe for revision with more sustained attention to the interaction between forms of orality and forms of literacy. It also speaks to the importance of non-alphabetic literacies, as Iroquois rituals, Dakota winter counts, and Pacific Northwest coast “totem poles” attest to other forms of record-keeping and a myriad of interactions that demonstrate cross-fertilization between oral and written forms. The colonizers also made deliberate use of cultural misrepresentation as a technique of coercion. Cessions of land, which the indigenous signatories thought of as temporary grants of use rights but which the English enforced as the entire alienation of all property rights, are well-known examples; for a survey, see Calloway, Pen and ink witchcraft ( 2013 ). For case studies, see Bross and Wyss, Early Native literacies ( 2008 ); Cohen, Networked wilderness ( 2010 ); Cushman, Cherokee syllabary ( 2011 ); S. Lyons, X-marks ( 2010 ); Morgan, Bearer of this letter ( 2009 ); Wyss, Writing Indians ( 2000 ).

For example, Canagarajah, Translingual practice ( 2013 ), repeats such catchwords and phrases as translingual, translocal, global, and cosmopolitan. Jan Blommaert’s 2008 pseudo-ethnography, Grassroots literacy , also slights these fundamental linguistic dimensions.

See also Cook-Gumperz and Gumperz, “From oral to written culture” ( 1981 ).

Among the large literature, see Schieffelin and Gilmore, Acquisition of literacy ( 1986 ); Tannen, Spoken and written language ( 1982 ); Heath, Ways with words ( 1983 ); Dyson, Multiple worlds of child writers ( 1989 ); Dyson, “‘Welcome to the jam’” ( 2003 ); Olson and Torrance, Cambridge handbook of literacy ( 2009 ).

For both examples of oppositions and differences, see Tannen, Spoken and written language ( 1982 ). The seminal work of Dell Hymes, Richard Bauman, and especially William Labov merits reopening by students of literacy.

See Street, Literacy in theory and practice ( 1984 ). For more on the debate over the New Literacy Studies, see Chap. 3 .

Bauman, Richard, and Charles L. Briggs. (1990) Poetics and performance as critical perspectives on language and social life. Annual Review of Anthropology 19: 59–88.

Google Scholar  

Bauman, Richard, and Joel Sherzer, eds. (1989) Explorations in the ethnography of speaking . 2d ed. Cambridge: Cambridge University Press.

Besnier, Niko. (1995) Literacy, emotion and authority: Reading and writing on a Polynesian atoll. Cambridge: Cambridge University Press.

Biber, Douglas. (2009). Are there linguistic consequences of literacy? Comparing the potentials of language use in speech and writing. In David R. Olson and Nancy Torrance, eds, The Cambridge handbook of literacy , 75–91. Cambridge: Cambridge University Press.

Blommaert, Jan. (2008) Grassroots literacy: Writing, identity, and voice in Central Africa . Milton Park, Abingdon, and New York: Routledge.

Boone, Elizabeth Hill, and Walter D. Mignolo, eds. (1994) Writing without words: Alternative literacies in Mesoamerica and the Andes . Durham: Duke University Press.

Brockmeier, Jens. (2000) Literacy as symbolic space. In Janet Wilde Astington, ed., Minds in the making: Essays in honor of David R. Olson , 43–61. Oxford: Blackwell.

Brockmeier, Jens. (2002) The literacy episteme: The rise and fall of a cultural discourse. In Literacy, narrative and culture , ed. Jens Brockmeier, Min Wang, and David R. Olson, 17–34. Richmond, England: Curzon.

Brockmeier, Jens, and David R. Olson. (2009) The literacy episteme: From Innis to Derrida. In David R. Olson and Nancy Torrance, eds., The Cambridge handbook of literacy , 3–21. Cambridge: Cambridge University Press.

Bross, Kristina, and Hilary E. Wyss, eds. (2008) Early Native literacies in New England: A documentary and critical anthology . Amherst: University of Massachusetts Press.

Calloway, Colin G. (2013) Pen and ink witchcraft: Treaties and treaty making in American Indian history . New York: Oxford University Press.

Canagarajah, Suresh. (2013) Translingual practice: Global Englishes and cosmopolitan relations. London: Routledge.

Cancian, Sonia. (2010) Families, lovers, and their letters: Italian postwar migration to Canada. Winnipeg: University of Manitoba Press.

Chafe, Wallace, and Deborah Tannen. (1987) The relation between written and spoken language. Annual Review of Anthropology 16: 383–407.

Clammer, J. R. (1976) Literacy and social change: A case study of Fiji . Leiden: E.J. Brill.

Clanchy, Michael T. (1993) From memory to written record: England, 1066–1307 . 2d ed. Oxford: Blackwell.

Cohen, Matt. (2010) The networked wilderness: Communicating in early New England. Minneapolis: University of Minnesota Press.

Cook-Gumperz, Jenny, and John J. Gumperz. (1981) From oral to written culture: The transition to literacy. In Marcia Farr Whitman, ed., Variation in writing: Functional and linguisticscultural differences , 89–109. Hillsdale, N.J.: Lawrence Erlbaum.

Cook-Gumperz, Jenny, ed. (1986) The social construction of literacy. Cambridge: Cambridge University Press.

Cushman, Ellen. (2011) The Cherokee syllabary: Writing the people’s perseverance . Norman: University of Oklahoma Press.

Dyson, Anne Haas. (1989) Multiple worlds of child writers: Friends learning to write . New York: Teachers College Press.

Dyson, Anne Haas. (2003) “Welcome to the jam: Popular culture, school literacy, and the making of childhoods. Harvard Educational Review 79: 328–361.

Finnegan, Ruth. (1973) Literacy versus non-literacy: The great divide. In Robin Horton and Ruth Finnegan, eds, Modes of Thought , 112–144. London: Faber and Faber.

Finnegan, Ruth. (1988) Literacy and Orality: Studies in the technology of communication . Oxford: Blackwell.

Gee, James Paul. (1986) Orality and literacy: From The Savage Mind to Ways with Words . TESOL Quarterly 20: 719–746.

Gee, James Paul. (1996) Social linguistics and literacies: Ideology in discourses . 2d ed. London: Routledge.

Goffman, Erving. (1981) Forms of talk. Philadelphia: University of Pennsylvania Press.

Goody, Jack. (1986) The logic of writing and the organization of society. Cambridge: Cambridge University Press.

Goody, Jack, and Ian Watt. (1968) The consequences of literacy. In Jack Goody, ed., Literacy in traditional societies , 27–68. Cambridge: Cambridge University Press.

Graff, Harvey J. (1979) The literacy myth: Literacy and social structure in the nineteenthcentury city . New York: Academic Press.

Graff, Harvey J. (1991) The literacy myth , reprinted with new introduction. New Brunswick, N.J.: Transaction.

Halverson, John. (1991) Olson on literacy. Language in Society 20: 619–640.

Halverson, John. (1992) Goody and the implosion of the literacy thesis. Man (new series) 27: 301–317.

Hanks, William F. (2010) Converting words: Maya in the age of the cross . Berkeley: University of California Press.

Harris, Roy. (2002) Literacy and the future of writing: An integrational perspective. In Jens Brockmeier, Min Wang, and David R. Olson, eds, Literacy, narrative and culture , 35–66. Richmond, England: Curzon.

Harris, William V. (1989) Ancient literacy . Cambridge, Mass: Harvard University Press.

Havelock, Eric A. (1963) Preface to Plato . Cambridge, Mass: Harvard University Press.

Havelock, Eric Alfred. (1973) Prologue to Greek literacy. Cincinnati: University of Cincinnati Taft Lectures.

Havelock, Eric Alfred. (1976) Origins of Western literacy. Toronto: Ontario Institute for Studies in Education.

Havelock, Eric A. (1977) The preliteracy of the Greeks. New Literary History 8: 369–392.

Havelock, Eric A. (1980) The coming of literate communication to Western culture. Journal of Communication 30: 90–99.

Havelock, Eric A. (1982) The literate revolution in Greece and its cultural consequences. Princeton: Princeton University Press.

Havelock, Eric A. (1986) The muse learns to write. New Haven: Yale University Press.

Heath, Shirley Brice. (1982) Protean shapes in literacy events: Ever-shifting oral and literate traditions. In Deborah Tannen, ed., Spoken and written language: Exploring orality and literacy , 91–117. Norwood, N.J.: Ablex.

Heath, Shirley Brice. (1983) Ways with words: Language, life, and work in communities and classrooms . Cambridge: Cambridge University Press; with new epilogue, 1996.

Heath, Shirley Brice. (1990) The children of Trackton’s children: Spoken and written language in social change. In James W. Stigler, Richard A. Shweder, and Gilbert S. Herdt, eds, Cultural psychology: Essays on comparative human development , 496–510. Cambridge: Cambridge University Press.

Heath, Shirley Brice. (2012) Words at work and play: Three decades in family and community life. Cambridge: Cambridge University Press.

Henkin, David. (2006) The postal age: The emergence of modern communications in nineteenthcentury America . Chicago: University of Chicago Press.

Hymes, Dell. (1962) The ethnography of speaking. In Thomas Gladwin and William C. Sturtevant, eds, Anthropology and human behavior , 15–53. Washington, D.C: Anthropological Society of Washington.

Hymes, Dell H. (1964a) A perspective for linguistic anthropology. In Sol Tax, ed., Horizons of Anthropology , 92–107. Chicago: Aldine.

Hymes, Dell H. (1964b) Introduction: Toward ethnographies of communication. American Anthropologist 66: 1–34.

Kalman, Judy. (1990) Writing on the plaza: Mediated literacy practice among scribes and clients in Mexico City . Cresskill, N.J: Hampton Press.

Koerner, Konrad. (1991) Toward a history of modern sociolinguistics. American Speech 66: 57–70.

Labov, William. (1972) Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.

Linell, Per. (2005) The written language bias in linguistics: Its nature, origins, and transformations. Milton Park, Abingdon, and New York: Routledge.

Lord, Albert B. (1960) The singer of tales . Cambridge, Mass.: Harvard University Press.

Lord, Albert B. (1995) The singer resumes the tale , ed. Mary Louise Lord. Ithaca: Cornell University Press, 1995.

Lord, Albert B. (2000) The singer of tales , 2d ed., ed. Stephen Mitchell and Gregory Nagy. Cambridge, Mass.: Harvard University Press.

Lurie, David B. (2011) Realms of literacy: Early Japan and the history of writing . Cambridge, Mass: Harvard University Asia Center.

Lyons, Martyn. (2001) Readers and society in nineteenth-century France: Workers, women, peasants . Basingstoke: Palgrave.

Lyons, Martyn. (2008) Reading culture and writing practices in nineteenth-century France . Toronto: University of Toronto Press.

Lyons, Martyn. (2013) The writing culture of ordinary people in Europe, c. 1860–1920. Cambridge: Cambridge University Press.

Lyons, Scott Richard. (2010) X-marks: Native signatures of assent. Minneapolis: University of Minnesota Press.

McKenzie, Donald F. (1999) Bibliography and the sociology of texts . Cambridge: Cambridge University Press.

Mignolo, Walter. (1995) The darker side of the Renaissance: Literacy, territoriality, and colonialization . Ann Arbor: University of Michigan Press.

Morgan, Mindy J. (2009) The bearer of this letter: Language ideologies, literacy practices, and the Fort Belknap community . Lincoln: University of Nebraska Press.

Olson, David R. (1994) The world on paper: The conceptual and cognitive implications of writing and reading. Cambridge: Cambridge University Press.

Olson, David R., and Nancy Torrance, eds. 2009. The Cambridge handbook of literacy . Cambridge: Cambridge University Press.

Ong, Walter J. (1958) Ramus, method, and the decay of dialogue. Cambridge, Mass.: Harvard University Press.

Ong, Walter J. (1967) The presence of the word. New Haven: Yale University Press.

Ong, Walter J. (1977) Interfaces of the word . Ithaca: Cornell University Press.

Ong, Walter J. (1982) Orality and literacy. London: Methuen.

Parry, Milman. (1971) The making of Homeric verse: The Collected Papers of Milman Parry , ed. Adam Parry. Oxford: Oxford University Press.

Rappaport, Joanne. (1990) The politics of memory: Native historical interpretation in the Colombian Andes . Durham: Duke University Press, 1990. 1

Rappaport, Joanne, and Tom Cummins. (2012) Beyond the lettered city: Indigenous literacies in the Andes . Durham: Duke University Press.

Romani, Gabriella. (2013) Postal culture: Writing and reading letters in post-unification Italy . Toronto: University of Toronto Press.

Säljö, Roger, ed. (1988) The written word: Studies in literate thought and action. Berlin: Springer-Verlag.

Salomon, Frank. (2001) How an Andean ‘writing without words’ works. Current Anthropology 42: 1–27.

Salomon, Frank, and Mercedes Niño-Murcia. (2011) The lettered mountain: A Peruvian village’s way with writing . Durham, N.C: Duke University Press.

Schieffelin, Bambi B., and Perry Gilmore, eds. (1986) The acquisition of literacy: Ethnographic perspectives. Norwood, N.J: Ablex.

Scribner, Sylvia, and Michael Cole. (1981) The psychology of literacy. Cambridge, Mass: Harvard University Press.

Seed, Patricia. (1991) “Failing to marvel”: Atahualpa’s encounter with the word. Latin American Research Review 26: 7–32.

Shuy, Roger W. (1990) A brief history of American sociolinguistics, 1949–1989. Historiographica Linguistica 17: 183–209.

Street, Brian. (1984) Literacy in theory and practice . Cambridge: Cambridge University Press.

Szwed, John F. (1981) The ethnography of literacy. In Marcia Farr Whiteman, ed., Variation in writing: Functional and linguistic-cultural differences , 13–23. Hillsdale, N.J.: Lawrence Erlbaum.

Tagupa, William E. H. (1981) Education, change, and assimilation in nineteenth century Hawai‘i. Pacific Studies 5: 57–69.

Tannen, Deborah, ed. (1982) Spoken and written language: Exploring orality and literacy . Norwood, N.J.: Ablex.

Thomas, Rosalind. (2009) The origins of western literacy: Literacy in ancient Greece and Rome. In David R. Olson and Nancy Torrance, eds., The Cambridge handbook of literacy , 346–361. Cambridge: Cambridge University Press.

Vincent, David. (1989) Literacy and popular culture: England 1750–1914. Cambridge: Cambridge University Press.

Vincent, David. (2000) The rise of mass literacy: Reading and writing in modern Europe. Cambridge: Polity Press.

Wolf, Maryanne. (2007) Proust and the squid: The story and science of the reading brain . New York: Harper.

Wyss, Hilary E. (2000) Writing Indians: Literacy, Christianity and Native community in early America . Amherst: University of Massachusetts Press.

Download references

Author information

Authors and affiliations.

Department of English, The Ohio State University, Columbus, OH, USA

Harvey J. Graff

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Harvey J. Graff .

Rights and permissions

Reprints and permissions

Copyright information

© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Graff, H.J. (2022). Linguistics: Between Orality and Writing. In: Searching for Literacy. Palgrave Macmillan, Cham. https://doi.org/10.1007/978-3-030-96981-3_2

Download citation

DOI : https://doi.org/10.1007/978-3-030-96981-3_2

Published : 22 August 2022

Publisher Name : Palgrave Macmillan, Cham

Print ISBN : 978-3-030-96980-6

Online ISBN : 978-3-030-96981-3

eBook Packages : Social Sciences Social Sciences (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Speech, Writing, and Sign

A Functional View of Linguistic Representation

Linguistics has traditionally dealt with questions about structure—what are the parts of a language and how are they assembled? Naomi Baron adopts a new approach by asking what a human language is used for and how it achieves its goals. She carefully examines what is communicated, why it is important. and how the exchange is accomplished. In the process of this basic redefinition, she fashions a lucid, systematic introduction to the study of linguistics. The initial chapters discuss language as a source and solution to problems of human communication, the various aspects of representation, the definition of human language, and a methodology for the functional analysis of language. The three chapters that follow fully explore this functional perspective for spoken, written, and signed languages, and offer new evidence to demonstrate the effect of social context on linguistic structure. Speech, Writing, and Sign is profusely illustrated with drawings, photographs, and reproductions of artistic examples. Written to be accessible to beginning students, this book will also interest linguistic scholars because of its challenges to current linguistic theory.

Table of Contents

• isbn 978-0-253-05120-2
• publisher Indiana University Press
• publisher place Bloomington, Indiana USA
• restrictions CC-BY-NC-ND
• rights Copyright © Trustees of Indiana University
• rights holder Indiana University Press
• rights territory World
• doi https://doi.org/10.2979/SpeechWritingandSign

We use cookies to analyze our traffic. Please decide if you are willing to accept cookies from our website. You can change this setting anytime in Privacy Settings .

This typology seem very rational, but in fact it is misleading, as rational taxonomies often are. All documented writing systems are a mixture of two or (usually) more of the these categories, and all include a significant phonological aspect. This critique has been most fully develop by John DeFrancis in his book Visible Speech , from which most of the examples below have been taken.

Given the definitions of writing we've given so far, pictographic and ideographic systems would not be included, since they are not ways of "recording language," but rather ways of directly picturing things, events and their relationships. Interestingly, as a matter of empirical fact, it seems that pictographic and ideographic systems have never really developed fully as such. This is not to say that people have never conveyed information with pictures, nor that sets of conventional icons standing for language-independent ideas have never been developed and used. In fact, pictographic and ideographic signs played a central role in the (various) inventions of writing.

However, pictographic or ideographic systems as such have never developed into a form fully capable of conveying unlimited messages from one person to another. Instead, they either remain as limited systems operating within a highly restricted application -- say to keep warehouse records -- or else they develop into a genuine writing system, capable of conveying any linguistic message. In the second case, the process of development into a genuine writing system always involves adding some phonological aspects, in ways we'll describe shortly.

Origins of writing

When they appear in the archeological record about 5,500 years ago, the Sumerians had developed a system of icons inscribed on clay tablets for keeping temple records. A typical example includes icons for "two", "sheep", "temple/house", and the gods "An" and "Inanna". The meaning might be "two sheep received from the temple of An and Inanna", or "two sheep delivered to the temple of An and Inanna", or perhaps something else entirely.

These marks constituted a limited notation system, which in the beginning may only have served to remind the writer of what he had once already known. However, as long as agreed-on standards were obeyed, another person could also read the record in the same way. In this, these were similar to systems for record-keeping, based on symbolic tokens of many sorts, developed over and over again in many cultures over the millennia -- marks on stone or bone, clay figurines, even knots in cords. As civilizations become more complex, record-keeping of this kind becomes increasingly important in order to keep commercial transactions straight. The ability of trained third parties to read such records in a consistent way became increasingly important as systems for mediating or adjudicating disputes in non-violent ways come into use. However, most such systems remained limited in their expressive capacity.

In the case of the Sumerian record-keeping system, two crucial innovations led (over a few hundred years) to a full writing system, capable of expressing anything that could be expressed in the (written) words of the Sumerian language.

The first innovation was the Rebus Principle : if you can't make a picture of something, use a picture of something with the same sound. The first clear example of this is in a tablet from Jemdet Nasr, dated to around 2900 BC, in which a pictograph of a reed ( GI in Sumerian) is used to mean "reimburse" (also pronounced GI ).

The second innovation was what we might call the Charades Principle : if you combine an ambigous or vague picture of the meaning of a word, with a little information about what the word sounds like, you can get a more effective communication of the identity of the word than if you tried to use only imperfect information about meaning, or imperfect information about sound. To give an example from Sumerian, a particular symbol having a meaning something like "leg" might be combined with a symbol pronounced "ba" to give the word GUB "to stand"; the same "leg" symbol, combined with a symbol pronounced "na", gave the word GIN "to go"; and combined with a symbol pronounced "ma", it gave the word TUM "to bring." Thus a Sumerian reader was in effect being asked to play a sort of game of charades : what word has something to do with "leg" and ends in the initial sound of "ba"? -- why of course, that's GUB, "to stand", what else! These combinations became conventionalized, resulting in a system that was presumably somewhat easier to learn to read than to learn to write, but was not very efficient in either direction.

Still, the result was a complete writing system, in which the Sumerians wrote down not just warehouse records, but poems, diplomatic treaties, letters, contracts and judicial decisions, dictionaries, and epic myths.

We can see a modern version of a similar system in Chinese characters. Most characters can be analyzed as containing two elements, one of which provides semantic information, while the other provides phonological information. The following small table (from DeFrancis) illustrates this with a set of four semantic elements crossed with a set of four phonological (or as DeFrancis calls it "phonetic") elements. The numbering of the semantic elements is taken from a standard set of 214 that have been recognized at least since the Kang Xi dictionary of the 18th century, while the numbering of the phonetic elements is taken from a list of 895 compiled by Soothill.

It is clearly inappropriate to call the Chinese system "ideographic", as is sometimes done. Chinese characters refer to morphemes, not ideas. However, to the extent that the pattern in the table above is taken as typical (and DeFrancis claims that about 75% of all Chinese characters work like these examples), Chinese characters are simultaneously a kind of syllabic writing. DeFrancis suggests the term "morpho-syllabic" to describe it.

It can be argued that the degree of phonological information found in the Chinese writing system is not radically different from what is found in English. English spelling usually tells us what the morphemes are, but unless we know in advance, it gives us only imperfect information about pronunciation. We can be sure that "tough" will not be pronounced "congressional" or "halter", but only knowledge of the word itself tells us that it rhymes with "rough" and not with "dough" or "through" or "plough".

Egyptian hieroglyphics also combined pictographic and phonological aspects, often in complicated ways, as the example below suggests. This is the word hememu "humanity". It starts with four symbols denoting the four consonants in the word (the symbol glossed with /u/ is actually /w/). It ends with three semantic determinatives: a seated man, a seated woman, and a set of three lines indicating that multiple entities are referenced.

This approach to writing produced a small number of symbols with simple phonetic values -- Egyptian had 24 simple consonant symbols, shown below -- and led naturally to the development of alphabetic writing systems.

Why pictographic/ideographic writing is not practical

No one has ever developed a full communications system based on pictographic or ideographic principles, although people have often surmised that this would be useful, because it would (or at least could) be universal. The problem is that universality means only that it is equally hard for everyone to develop and learn such a system. If it is feasible to design such a system at all, it is at least very, very difficult. Since everyone already knows at least one ordinary spoken language, practical people will always tend to give up on the ideographic system and start using a written form of their speech, as soon as they can figure out how to do this.

For an amusing myth about this process, check out the story of How the first letter was written , from Rudyard Kipling's Just So Stories .

Why phonological writing is (eventually) practical

It is rather difficult to get enough conscious access to the phonological structure of speech to design an alphabetic writing system, and very few languages have small enough inventories of syllables for a syllabic system to be an easy place to start. More important, the idea of constructing a full writing system (on any basis, phonological or otherwise) is not at all an obvious one.

So writing seems to have started with pictograms for mnemonic aids in record keeping, or as vehicles of insight in divination. As the inventory of signs increases, the possibility arises to begin using some of the signs as rebuses or as phonological/semantic combinations. This is much more efficient than trying to design a new symbol for every word or morpheme. Once this meaning-plus-sound process begins, it can develop into a full (if complex and inefficient) writing system, able to encode any passage in the language. This development seems to have occurred independently at least three times: in the middle east; in China; and in Mexico.

Various other developments are then logically possible. The Chinese (and other cultures influenced by them, including Japan) developed a meaning-plus-sound system based on the syllabic unit. The Mayans did the same. A logical next step is to increase efficiency by doing away with some or all of the meaning-related units, in favor of a consistent syllabary of some sort. Such syllabaries were developed throughout the far east, but in most cases they did not displace the mean-plus-sound elements. Instead they supplemented them for certain uses (such as the encoding of grammatical particles in Japanese) or for certain populations (such as women in some places and periods in China).

By contrast, the Egyptians (and other semitic languages) developed a meaning-plus-sound system based on primarily or solely on consonants. This naturally led to purely consonantal writing systems for some of the semitic languages (such as Phonecian), which shared with Egyptian the property of changing vowels extensively for morphological purposes. To give an example from Hebrew, the root /ktb/ can have among many other forms katav "I wrote", kotav "I write", katoov "written", kitav "letters", katban "scribe". In a language that works this way, it's natural to factor words into consonants and vowels, and to start with a sort of acronym-like use of pictographs to denote their initital consonant, in a meaning-plus-sound system based on consonants only. In the case of most if not all Semitic languages, it has turned out to be usually possible to figure out the vowels from context, even without adding semantic determinatives. This made it possible to abandon the semantic determinatives without giving up general writing.

Alphabetic systems seem to be rather unnatural, and have arguably been developed only once, by the Greeks when they adapted the Semitic consonant-only system to their language, which couldn't so easily be written without vowels. It is possible that this invention would not have happened at all without this particular historical sequence.

Part of the reason for the success of meaning-plus-sound systems is that two kinds of evidence are always better than one: two fairly lousy systems can be combined into one decent one. But there is another reason as well. Sound systems are made up of quite limited materials. In many lanugages, the number of distinct syllables is not terribly great; and the number is much reduced if natural equivalence classes (such as "starts with" or "rhymes with") are used. Once one gets started down the sound-system road, it is tempting to go all the way, since it makes the practical training of scribes more efficient only to have to learn a few dozen symbols, rather than several thousand. Of course, the Scribe's Guild may think it's just fine to limit the number of literate people, and to leave large barriers in place, blocking entry to skill in their profession.

Why is reading hard to learn?

For the same reasons that writing was hard to invent, reading is hard to learn. Neither reading nor writing is a biologically natural process. Alphabetic writing systems are in principle the most efficient, since they require learning the smallest number of symbols. No one would design a writing system today on any other basis. However, alphabetic systems seem to impose a special burden on learners, because they require understanding a level of analyses -- phonemic analysis -- that is relatively inaccessible to introspective scrutiny. The orthographic system of English also has many morpheme-related idiosyncrasies, which eventually make it easier to recognize words (just as the Chinese morphosyllabic system does), but which also may obscure the alphabetic principle for early learners.

A number of ongoing long-term longitudinal studies show that about 60% of American children find it difficult to learn to read, and that 20-30% fall seriously behind or fail entirely. The reasons for these problems, and the best ways to deal with them, are a matter of great controversy. A great deal depends on the answers.

The opening salvo in one of the this war's battles was fired more than 50 years ago by Rudolf Flesch in his 1955 book Why Johnny Can't Read .

In an attempt to present an authoritative consensus on this important topic, I'll start with quotes from congressional testimony given in 1998 and 1999 by Dr. G. Reid Lyon, Chief of the Child Development and Behavior Branch of the National Institute of Child Health and Human Development, which in turn is part of the National Institututes of Health (NIH).

First, the diagnosis:

[C]hildren who have difficulties learning to read can be readily observed. The signs of such difficulty are a labored approach to decoding or "sounding" unknown or unfamiliar words and repeated misidentification of known words. Reading is hesitant and characterized by frequent starts and stops and multiple mispronunciations. If asked about the meaning of what has been read, the child frequently has little to say. Not because he or she is not smart enough; in fact, many youngsters who have difficulty learning to read are bright and motivated to learn to read, at least initially. Their poor comprehension occurs because they take far too long to read the words, leaving little energy for remembering and understanding what they have read. Unfortunately, there is no way to bypass this decoding and word recognition stage of reading. Using context to figure out the pronunciation of unknown words cannot appreciably offset a deficiency in these skills. In essence, while one learns to read for the fundamental purpose of deriving meaning from print, the key to comprehension starts with the immediate and accurate reading of words. In fact, difficulties in decoding and word recognition are at the core of most reading difficulties... To be sure, there are some children who can read words accurately and quickly yet do have difficulties comprehending, but they constitute a small portion of those with reading problems. If the ability to gain meaning from print is dependent upon fast, accurate, and automatic decoding and word recognition, what factors hinder the acquisition of these basic reading skills? As mentioned above, young children who have a limited exposure to both oral language and print before they enter school are at-risk for reading failure. However, many children with robust oral language experience, average to above intelligence and frequent interactions with books since infancy may also show surprising difficulties learning to read. Why? In contrast to good readers who understand that segmented units of speech can be linked to letters and letter patterns, poor readers have substantial difficulty in developing this "alphabetic principle." The culprit appears to be a deficit in phoneme awareness--the understanding that words are made up of sound segments called phonemes. Difficulties in developing phoneme awareness can have genetic and neurobiological origins or can be attributable to a lack of exposure to language patterns and usage during the preschool years. The end result is the same, however. Children who lack phoneme awareness have difficulties linking speech sounds to letters, leading to limitations in the development of decoding and word recognition skills, resulting in extremely slow reading. As mentioned, this inaccurate and labored access to print renders comprehension very difficult.

It is clear from research on emerging literacy that learning to read is a relatively lengthy process that begins very early in development and clearly before children enter formal schooling. Children who receive stimulating literacy experiences from birth onward appear to have an edge when it comes to vocabulary development, an understanding of the goals of reading, and an awareness of print and literacy concepts. Children who are read to frequently at very young ages become exposed in interesting and exciting ways to the sounds of our language, to the concept of rhyming, and to other word and language play that serves to provide the foundation for the development of phoneme awareness. As children are exposed to literacy activities at young ages, they begin to recognize and discriminate letters. Without a doubt, children who have learned to recognize and print most letters as preschoolers will have less to learn upon school entry. The learning of letter names is also important because the names of many letters contained the sounds they most often represent, thus orienting youngsters early to the alphabetic principle or how letters and sounds connect. Ultimately, children's ability to understand what they are reading is inextricably linked to their background knowledge. Very young children who are provided opportunities to learn, think, and talk about new areas of knowledge will gain much from the reading process. With understanding comes the clear desire to read more and to read frequently, ensuring that reading practice takes place.

In studying approximately 34,501 thousand children over the past 33 years, we have learned the following with respect to the role that phonemic awareness plays in the development of phonics skills and fluent and automatic word reading: Phonemic awareness skills assessed in kindergarten and first grade serve as potent predictors of difficulties learning to read. We have learned how to measure phonemic awareness skills as early as the first semester in kindergarten with tasks that take only 15 minutes to administer - and over the past decade we have refined these tasks so that we can predict with approximately 80% to 90% accuracy who become good readers and who will have difficulties learning to read. We have learned that the development of phonemic awareness is a necessary but not sufficient condition for learning to read. A child must integrate phonemic skills into the learning of phonics principles, must practice reading so that word recognition becomes rapid and accurate, and must learn how to actively use comprehension strategies to enhance meaning. We have begun to understand how genetics are involved in learning to read, and this knowledge may ultimately contribute to our prevention efforts through the assessment of family reading histories. We are entering very exciting frontiers in understanding how early brain development can provide a window on how reading develops. Likewise, we are conducting studies to help us understand how specific teaching methods change reading behavior and how the brain changes as reading develops. We have learned that just as many girls as boys have difficulties learning to read. Until five years ago, the conventional wisdom was that many more boys than girls had such difficulties. Now females should have equal access to screening and intervention programs. We have learned that for 90% to 95% of poor readers, prevention and early intervention programs that combine instruction in phoneme awareness, phonics, fluency development, and reading comprehension strategies, provided by well trained teachers, can increase reading skills to average reading levels. However, we have also learned that if we delay intervention until nine-years-of-age, (the time that most children with reading difficulties receive services), approximately 75% of the children will continue to have difficulties learning to read throughout high school. To be clear, while older children and adults can be taught to read, the time and expense of doing so is enormous.

For a more recent summary of the same issues, see Mark Seidenberg , " The Science of Reading and its Educational Implications ", 2013:

Research in cognitive science and neuroscience has made enormous progress toward understanding skilled reading, the acquisition of reading skill, the brain bases of reading, the causes of developmental reading impairments and how such impairments can be treated. My question is: if the science is so good, why do so many people read so poorly? I mainly focus on the United States, which fares poorly on cross-national comparisons of literacy, with about 25-30% of the population exhibiting literacy skills that are low by standard metrics. I consider three possible contributing factors, all of which turn on issues concerning the relationships between written and spoken language. They are: the fact that English has a deep alphabetic orthography; how reading is taught; and the impact of linguistic variability as manifested in the Black-White “achievement gap”. I conclude that there are opportunities to increase literacy levels by making better use of what we have learned about reading and language, but also institutional obstacles and understudied issues for which more evidence is badly needed.

This may sound like a case where common sense is confirmed by the results of scientific study. However, the implications cannot be taken for granted. The " Whole Language " approach, which at times has dominated American educational practice, is still strongly represented among teachers and educational administrators. This approach emphasizes a direct between seeing written words and understanding their meaning, featuring written words as visual patterns, and avoiding any focus on the systematic relationship between letters and sounds. Despite the evident good intentions of its adherents, the Whole Language movement has been a deeply destructive force.

See this 2018 Forbes Magazine article " Why Johnny Still Can't Read -- And What To Do About It ", and this review of Mark Seidenberg's 2017 book Language at the Speed of Sight for a recent evaluation of the situation, which seems to be that most American teachers are taught little or nothing about the science of reading and how to teach reading effectively.

The result is that many American children are still not learning to read well. Thus according to " Reading Scores on National Exam Decline in Half the States ", NYT 10/30/2019:

Only 35 percent of fourth graders were proficient in reading in 2019, down from 37 percent in 2017; 34 percent of eighth graders were proficient in reading, down from 36 percent.

This debate about how to teach children to read is one of the most important public policy issues in the country today.

• Search Menu
• Sign in through your institution
• Advance articles
• Author Guidelines
• Submission Site
• Open Access
• Why Submit?
• About Applied Linguistics
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Terms and Conditions
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Introduction, reflexivity of metadiscourse and metadiscourse markers in the corpus, methodology, findings and discussion, summary and conclusion, supplementary data, acknowledgements.

• < Previous

Variation in Metadiscourse across Speech and Writing: A Multidimensional Study

• Article contents
• Figures & tables
• Supplementary Data

Man Zhang, Variation in Metadiscourse across Speech and Writing: A Multidimensional Study, Applied Linguistics , Volume 43, Issue 5, October 2022, Pages 912–933, https://doi.org/10.1093/applin/amac012

• Permissions Icon Permissions

Until now, there has been very little unified analysis of metadiscourse across speech and writing. Drawing on a reflexive metadiscourse model, this article conducts a multidimensional analysis of metadiscourse across 10 spoken and written registers in a corpus of 626 texts. Two metadiscourse dimensions, participants’ interaction and discourse presentation, are identified. The more interactive registers are more prominently marked by participants’ interaction, and the more informational registers by discourse presentation. Along metadiscourse dimensions, the 10 registers are lumped into 6 clusters: ‘discussion broadcasts and public conversations’, ‘scripted speeches and unscripted speeches’, ‘general prose and nondiscussion broadcasts’, ‘press and fiction’, ‘casual conversations’, and ‘academic prose’. Metadiscourse markers of both dimensions are frequently used in registers of the first two clusters which are highly interactive and informational, while they rarely occur in registers of the next two clusters which have low interactive or informational focuses. Metadiscourse markers of participants’ interaction are pervasive in the highly interactive register (casual conversations), and those of discourse presentation in the highly informational register (academic prose).

Metadiscourse in various registers, especially academic registers and written registers, has attracted great scholarly interest over the last three decades. Until now, metadiscourse has been investigated in research articles (e.g. Mauranen 1993a ; Jiang and Hyland 2016 ), students’ writings (e.g. Crismore et al. 1993 ; Ädel 2006 ), textbooks (e.g. Crismore 1989 ), teachers’ feedback (e.g. Ädel 2017 ), book reviews (e.g. Tse and Hyland 2008 ), lectures (e.g. Bouziri 2021 ), seminars (e.g. Mauranen 2010 ), journalistic registers (e.g. Khabbazi-Oskouei 2013 ), business writing (e.g. Hyland 1998 ), and so on. These studies reveal that metadiscourse is highly register-specific. Strong patterns of metadiscourse use in one register often represent weak patterns in another register. Surprisingly, despite the increasing research interest in various registers and the widely acknowledged importance of register variation, few studies have compared registers, and there has been very little attempt to provide a unified analysis of metadiscourse across spoken and written registers. 1 However, only by comparison can we determine whether a given pattern is notably common or rare in a register, and the analysis of one or two registers is not adequate for conclusions concerning metadiscourse use as a whole.

This research gap is at least partly due to methodological difficulties. A register is a discourse type associated with a particular situation of use ( Halliday and Hasan 1989 : 38–39; Biber and Conrad 2009 : 20). All metadiscourse markers of a register can be said to be used in association with the situational context. However, it is difficult, if not impossible, to analyze all metadiscourse markers of all registers. Besides, it is unreliable to compare registers in terms of individual metadiscourse markers because of their disparity across texts. We have tried to fill this gap, and investigated metadiscourse across written registers ( Zhang 2016 ) and spoken registers ( Zhang et al. 2017 ) resorting to the multidimensional framework. Multidimensional analysis is a corpus-based computational method for the comprehensive analysis of register variation ( Biber 1988 , 1995 ). It uses multivariate statistical techniques to identify dimensions of variation underlying metadiscourse, which reduces the large number of metadiscursive expressions to a small set of underlying dimensions and enables the description of relations among various registers in the space defined by the dimensions. We identified three written metadiscourse dimensions (writer presence, text presentation, reader guidance) ( Zhang 2016 ) which are quite different from the three spoken metadiscourse dimensions (two-way communication, united organization, discourse presentation) ( Zhang et al. 2017 ). These findings further prove that metadiscourse use varies across registers, and the overall description of English metadiscourse requires a unified investigation of metadiscourse across spoken and written registers.

Metadiscourse is reflexive, and metadiscourse markers refer reflexively to the discourse and its producer and recipient. This study, using the multidimensional approach and the corpora of the British National Corpus (BNC) and the Freiburg Lancaster–Oslo–Bergen Corpus (F-LOB), analyzes and compares the reflexive metadiscourse markers in 10 spoken and written registers: nondiscussion broadcasts, discussion broadcasts, scripted speeches, unscripted speeches, public conversations, casual conversations, press, general prose, academic prose, and fiction. It mainly answers two research questions: (i) what are the basic dimensions of variation in English metadiscourse? (ii) in what way are spoken and written registers different or similar along the dimensions? In the following, this article first defines metadiscourse, and classifies metadiscourse markers found in the corpus according to the definition. Second, it describes the corpus and analytical steps of the multidimensional analysis of metadiscourse. Third, it interprets the two basic dimensions of variation in English metadiscourse identified by a factor analysis, compares the 10 spoken and written registers along the two metadiscourse dimensions, and develops a typology of registers in terms of metadiscourse using a cluster analysis. The final section summarizes and concludes the whole article.

Reflexivity is the capacity of natural languages for referring to or describing themselves ( Lyons 1977 : 5). This important feature of human language makes human communication unique and different from that of honey bees, chimpanzees, and so on ( Sinclair 2005 : 164). Metadiscourse is reflexive in the sense that it talks about the ongoing language event or the discourse itself. More specifically, metadiscursive expressions refer reflexively to the discourse and its addresser and addressee (e.g. Williams 1981 : 235; Vande Kopple 1985 : 83; Crismore et al. 1993 : 39–40; Hyland 2005 : 14; Ädel 2006 : 20; Aguilar 2008 : 17). For example, This article focuses on … talks about the discourse, I want to talk about … the addresser of the discourse, and You’ll find … the addressee of the discourse.

From the first use of the term ‘metadiscourse’ ( Harris 1959 ) and especially since the 1980s when metadiscourse was adopted in discourse studies (e.g. Williams 1981 ; Vande Kopple 1985 ; Crismore 1989 ), reflexivity has been taken as one defining feature of metadiscourse. It has been emphasized in a variety of definitions of metadiscourse, such as ‘writing about writing’ ( Williams 1981 : 211–212), ‘discourse about discourse’ ( Vande Kopple 1985 : 83; Crismore et al. 1993 : 40), ‘text about text’ ( Mauranen 1993a : 4), ‘self-reflective expressions’ ( Hyland 2005 : 37), and ‘text about the evolving text’ ( Ädel 2006 : 20).

(1) Let me give you one or two specific examples […] (F8B) 2 (2) […] when you cut the palm of your hand there is a lot of blood pumping out […] (F8D)

Following Zhang (2016) and Zhang et al. (2017) , this research uses metadiscourse markers in the form of words and phrases as the analytical unit (see Supplementary Appendix Table A ), and metadiscourse markers in the corpus (see ‘Corpus’ section) are classified into three broad categories according to their functions: those referring to speech event components, describing discourse actions, and describing discourse circumstances (see Table 1 and Supplementary Appendix Table A ). Metadiscourse markers referring to speech event components are nouns and pronouns. They refer to the discourse (e.g. article ), addresser (e.g. I ), addressee (e.g. you ), and addresser and addressee (e.g. we ). Discourse actions described by verbal metadiscourse markers are of five types: introducing and analyzing (e.g. explore ), planning and arranging (e.g. arrange ), saying and arguing (e.g. argue ), defining and explaining (e.g. define ), and receiving and understanding (e.g. hear ). Circumstances concern the ‘when, where, how, to what extent’ in the discourse ( Biber et al. 1999 : 763). In the corpus, phoric markers (e.g. above ), code gloss markers (e.g. namely ), and style markers (e.g. briefly ) describe discourse circumstances, and they are typically adverbials. These metadiscourse markers are linguistic features for the multidimensional analysis of this research (see ‘Analytical steps’ section). Using smaller chunks of discourse (words and phrases) as the unit of analysis makes computer-assisted methods more applicable.

Metadiscourse markers in the corpus

Metadiscourse is a contextual phenomenon ( Mauranen 1993b : 173; Hyland 2005 : 24–25; Ädel 2006 : 25–26). Metadiscourse markers can only be identified in the context. This sets strict limits on the corpus for the comprehensive metadiscourse research across speech and writing. First, the corpus should be large enough to contain a wide range of registers. Second, the corpus should be small enough for the manual analysis. Third, all texts of the corpus should be available for metadiscourse analysis in the context. After comparing many corpora, I decided on BNC for the spoken registers and F-LOB for the written registers. Actually, they were the only corpora meeting the above requirements that I had access to although they are relatively dated. Table 2 lists the 10 registers investigated in the present analysis: 6 spoken registers from BNC and 4 written registers from F-LOB. Altogether, there are 626 2,000-word text samples totaling 1,252,000 words.

Registers in the corpus

BNC which was compiled in the early 1990s is a general corpus of British English. It is a 100-million word corpus containing both written (90 per cent) 3 and spoken (10 per cent) texts. This research analyzes metadiscourse in the six spoken registers of BNC: nondiscussion broadcasts, discussion broadcasts, scripted speeches, unscripted speeches, public conversations, and casual conversations ( Lee 2001 ), among which nondiscussion broadcasts, scripted speeches, and unscripted speeches are generally monologues, and discussion broadcasts, public conversations, and casual conversations are generally dialogues. Each of the six broad registers includes sub-registers which differ by participants, channels, settings, communicative purposes, topics, and so on (see Table 2 ).

As the spoken part of BNC is too large for the manual analysis, I used purposive, stratified, and random sampling methods to select some texts of different sub-registers for each of the six broad registers. The sampling process was not straightforward because sub-registers under different broad registers and texts of different sub-registers vary in number widely. The general principles of sampling are that texts of each broad register should cover as many sub-registers as possible, texts of different sub-registers in the same broad register should not vary greatly in number, and text numbers of all six broad registers should be the same. Taking into consideration these three principles, I selected 21 texts for each broad register (see Table 2 ). As some texts are too long for the present research (more than 100,000 words), I randomly chose about 2,000 words from each text which is the size preferred by F-LOB. This size has also been used with success by many other corpora (e.g. corpora of the Brown family and the International Corpus of English). If a text selected had less than 2,000 words, I would repeat the procedures to select another text. In all random sampling cases, it was done at a sentence boundary.

F-LOB is also a general corpus of British English built in the early 1990s. This one-million-word corpus comprises 500 written text samples (about 2,000 words per text) covering 4 broad registers (press, general prose, academic prose, 4 and fiction) and 15 sub-registers (see Table 2 ). The whole corpus of F-LOB is used, and the four broad registers are compared in the present research.

Analytical steps

In this multidimensional study, metadiscourse dimensions are identified using factor analysis, registers are compared along the dimensions, and registers that are similar in terms of the dimensions are grouped together through cluster analysis. 5 The multidimensional approach uses factor analysis to discover co-occurrence relations among linguistic features ( Biber 1988 ). In this study, the linguistic features examined are metadiscourse markers. A metadiscourse factor represents a grouping of metadiscourse markers that consistently co-occur in texts. Linguistic features on the same factor perform the same situational, social, or cognitive function ( Biber 1988 : 101). Metadiscourse markers co-occur systematically because they are used for a shared function. Based on the shared function, a metadiscourse factor is labeled functionally. A labeled metadiscourse factor is called a metadiscourse dimension. A metadiscourse dimension represents a group of co-occurring metadiscourse markers which share a situational, social, or cognitive function ( Zhang 2016 : 211). Metadiscourse dimensions are parameters of variation in metadiscourse. Having identified metadiscourse dimensions, the similarities and differences among registers in metadiscourse can be considered. However, the multidimensional description of the overall relations among registers is not straightforward because there is more than one dimension to consider and the same register may have idiosyncratic distributions along different dimensions. In order to achieve both descriptive adequacy and simplicity, a cluster analysis (e.g. Everitt et al. 2011 ; Moisl 2015 ) is conducted to divide the 10 registers into homogeneous groups.

The multidimensional analysis follows five methodological steps: building a corpus, identifying metadiscourse markers, extracting metadiscourse dimensions, describing register variation, and developing a register typology. This research uses a collection of 626 texts of 10 spoken and written registers taken from BNC and F-LOB (see ‘Corpus’ section). In these texts, reflexive metadiscourse markers are identified manually, which is very time- and energy-consuming. Potential metadiscourse markers were found by surveying previous research and reading part of the corpus manually. In the corpus, every 10th of the 500 texts in F-LOB and 500 words randomly selected from each spoken text in the BNC Sampler built to mirror the full BNC were read to locate potential metadiscourse markers. They are ‘potential’ metadiscourse markers because they may or may not be used metadiscursively, and can only be identified as metadiscursive or not in the context. All occurrences of potential metadiscourse markers were then retrieved from the corpus by Antconc 3.2.4w ( Anthony 2013 ) automatically. They were analyzed manually in the context to determine whether they are used metadiscursively in different contexts, and metadiscourse markers were classified according to their functions (see Table 1 ). Altogether, 271 potential metadiscourse markers were found in the corpus (see Supplementary Appendix Table A ). These expressions are used metadiscursively at least once in the corpus.

Seven analysts (including the writer of this article) who were English teachers or postgraduates majoring in English joined the manual analysis. The writer trained the rest analysts to identify and classify a metadiscourse marker, and all analysts agreed on the criterion of metadiscourse identification (i.e. reflexivity) and the classification of metadiscourse markers (see Table 1 ). Each case was analyzed by at least two analysts separately, and they agreed in more than 90 per cent of the cases. When disagreement occurred, analysts discussed to reach agreement. The writer joined, monitored and supervised the whole process, and read over all the results. The manual analysis identified metadiscourse markers, and at the same time established their frequencies which were basis for the factor analysis.

The software SPSS (19.0) was used for the factor analysis which identified factorial structures underlying metadiscourse markers. Following Biber (1988) , I used Principal Axis Factoring with Promax rotation to extract factors, and the scree plot to determine the best number of factors. The preliminary data screening, Bartlett’s sphericity test (approx. x 2 = 4,546.599, df = 66; p < 0.0001) and the Kaiser–Meyer–Olkin (KMO) index (KMO = 0.760) proved that our data are suitable for factor analysis. The scree plot of eigenvalues (see Figure 1 ) showed a clear break between the second and third factors, which means additional factors from Factor 3 contribute little to the overall analysis. The first two factors that account for 60.324 per cent of the shared variance were retained. They seem to be well-defined theoretically (see ‘Findings and Discussion’ section), which further supports this two-factor solution. The factorial structure underlying metadiscourse markers is given in Table 3 . Two metadiscourse factors were interpretively labeled ‘participants’ interaction’ and ‘discourse presentation’ (see ‘FINDINGS AND DISCUSSION’ section for the interpretation).

Scree plot of eigenvalues.

Two metadiscourse dimensions

The decimal numbers in Table 3 are factor loadings or weights. Factor loadings indicate ‘the degree to which one can generalize from a given factor to an individual linguistic feature’ ( Biber 1988 : 81). The factor loadings of metadiscourse markers on a metadiscourse factor show the degree to which these metadiscourse markers represent this factor. Metadiscourse markers with higher loadings are better representatives of a factor. In labeling metadiscourse factors, functions of metadiscourse markers with higher loadings are given priority. On each factor, only metadiscourse markers with loadings higher than |0.30| (absolute values) are included because the minimum loading weight that is considered significant is |0.30| ( Costello and Osborne 2005 : 4). If a feature has salient loadings on more than one factor, it is only considered on the factor where it has the highest loading to ‘assure the experimental independence of the factor scores’ ( Biber 1988 : 93). For example, metadiscourse markers referring to the addresser and addressee has a loading of 0.316 on Metadiscourse Factor 1, and 0.380 on Metadiscourse Factor 2, and they are only included as a feature on Factor 2. In this article, negative loadings are not analyzed because their absolute values are lower than 0.30.

Having identified two metadiscourse dimensions, the distribution of the 10 registers along each dimension and the overall relations among registers along both dimensions can be described with respect to their mean dimension scores. The factor analysis using SPSS computed a dimension score for each text along each dimension. The mean dimension score of a register along a metadiscourse dimension can be computed by dividing the sum of dimension scores of all texts of this register along this dimension by the number of texts of this register. Standard deviations measure the spread of distribution. The standard deviations of the dimension scores for a register show how tightly most texts in the register are grouped around the mean score along a dimension. A General Linear Models procedure and Tukey’s HSD post hoc tests were performed to test for the statistical significance of differences among registers in metadiscourse based on dimension scores. A p <0.05 is regarded as statistically significant, a p <0.001 statistically highly significant, and a p >0.05 not significant. R 2 scores represent the percentage of variance in the dimension score that can be predicted by knowing the register distinctions. They indicate the importance of a dimension in distinguishing among the registers. Mean scores, standard deviations, p -values, and R 2 scores are presented in Figures 2 and 3 (see ‘FINDINGS AND DISCUSSION’ section for the interpretation).

Mean dimension scores of the 10 registers along Metadiscourse Dimension 1 (ANOVA, F = 167.151, p = 0.000 < 0.001, R 2 = 70.9 per cent).

Note : The two figures outside and inside the brackets are mean scores and standard deviations, respectively. acad pro: academic prose; gen pro: general prose; prss: press; fict: fiction; nondis br: nondiscussion broadcasts; dis br: discussion broadcasts; scr spch: scripted speeches; unscr spch: unscripted speeches; pub conv: public conversations; cas conv: casual conversations .

Mean dimension scores of the 10 registers along Metadiscourse Dimension 2 (ANOVA, F = 46.906, p = 0.000 < 0.001, R 2 = 40.7 per cent). Note : The two figures outside and inside the brackets are mean scores and standard deviations, respectively. acad pro: academic prose; gen pro: general prose; prss: press; fict: fiction; nondis br: nondiscussion broadcasts; dis br: discussion broadcasts; scr spch: scripted speeches; unscr spch: unscripted speeches; pub conv: public conversations; cas conv: casual conversations.

Based on mean dimension scores of the 10 registers on both metadiscourse dimensions, a cluster analysis is performed to cluster together registers that are maximally similar in metadiscourse. The mean dimension scores were entered into a hierarchical cluster analysis (SPSS 19.0). I used the method of the furthest neighbor which produces compact clusters and the measure of the squared Euclidean distance which is the default option of SPSS. To display the clustering procedure and output clearly and vividly, the proximity matrix, agglomeration schedule, and dendrogram were selected. The cluster analysis identified six clusters of registers (see ‘Register clustering in terms of metadiscourse dimensions’ section).

Dimensions of variation in metadiscourse

(3) […] I can tell you that it beats jogging out of sight. (E28) (4) Oh do speak up I can’t hear you. (HV0) (5) What do you mean they’ve no choice? I mean we’re not just this with cats, […] (HV3)

In Example (3), the metadiscourse marker I refers to the writer, you refers to the reader, and tell describes the action of saying and arguing. They make it explicit that the writer is expressing something to the reader. In Examples (4) and (5), I is a metadiscourse marker of addresser, you addressee, speak saying and arguing, hear receiving and understanding, and mean defining and explaining. They facilitate the interaction between discourse participants, and help to increase their mutual understanding.

(6) […] this will be dealt with in more detail when the case study is discussed later up the paper. (J72) (7) For example, consider his case of a contemporary of ours […] (G63) (8) Shall we now move on to er the next paper […] (FUM).

Example (6) introduces the topic (‘this’) of a certain part of the text (‘later up the paper’) using metadiscourse markers of discourse ( paper ) and introducing and analyzing ( deal with and discuss ), a phoric marker ( later ) and a style marker ( in more detail ). Example (7) uses a code gloss marker ( for example ) and a metadiscourse marker of introducing and analyzing ( consider ) to offer some explanation. In Example (8), we refers to the addresser and addressee, move describes the discourse action of planning and arranging, and now and next are phoric markers. These metadiscourse markers are used to organize the text.

(9) We begin by reviewing the relevant background material. (J21)

In addition to the two-dimensional solution, eigenvalues from the factor analysis indicate the amount of variance accounted for by each dimension. In this study, Metadiscourse Dimension 1 accounts for 36.765 per cent of the shared variance, and Metadiscourse Dimension 2 an additional 23.559 per cent. It shows that both dimensions are significant. Dimension 1 accounts for greater proportion of variance and is the more basic and powerful dimension of variation among registers.

The present model establishes two basic dimensions of variation underlying metadiscourse: participants’ interaction and discourse presentation. It is similar to previous metadiscourse models (e.g. Vande Kopple 1985 ; Crismore et al. 1993 ; Hyland 2005 ; Ädel 2006 ) to various degrees in that one dimension is oriented toward the discourse participants, and one toward the discourse. There are, however, two major differences between this and earlier models. First, in this model, linguistic dimensions rather than functional dimensions are given priority. This research begins with identifying groups of co-occurring metadiscourse markers based on their frequencies, and these groups are interpreted functionally afterward. With the help of quantitative statistical techniques, this model recognizes two independent metadiscourse dimensions, and each dimension consists of an independent group of co-occurring metadiscourse markers. Besides, the importance of metadiscourse dimensions and metadiscourse markers on each dimension are defined precisely by eigenvalues and factor loadings. Previous models (e.g. Vande Kopple 1985 ; Crismore et al. 1993 ; Hyland 2005 ; Ädel 2006 ) also identify sets of metadiscourse markers. But the functional distinction is the basis of the grouping. It is not checked whether or to what extent these sets are separate dimensions, and it is often complained that there is overlap between different sets (e.g. Crismore 1989 : 65; Ädel 2006 : 23; Hyland 2017 : 18). Our research (including Zhang 2016 ; Zhang et al. 2017 ) tries to solve this problem. Take, for example, action markers or illocution markers (e.g. ‘to sum up’ and ‘I argue that’). They are usually classified as textual or discourse-oriented metadiscourse (e.g. Vande Kopple 1985 : 84; Ädel 2006 : 98). This study discovers that different action markers perform different functions in texts. Those of saying and arguing, receiving and understanding, and defining and explaining describe the interaction between discourse participants, while those of introducing topic and analyzing and planning and arranging introduce the discourse.

Second, this model describes functional dimensions of metadiscourse variation among registers more generally and comprehensively. It is built on consideration of 10 common types of speech and writing, while earlier models are based on metadiscourse analysis in much more restricted registers. Comparing, for example, metadiscourse dimensions in this model with metadiscourse dimensions extracted in writing (writer presence, text presentation, reader guidance) ( Zhang 2016 ) and speech (two-way communication, united organization, discourse presentation) ( Zhang et al. 2017 ), both metadiscourse dimensions and metadiscourse markers along each dimension are different. Roughly speaking, written metadiscourse dimensions of writer presence and reader guidance are combined into that of participants’ interaction in this study, and spoken metadiscourse dimensions of united organization and discourse presentation into that of discourse presentation. As for metadiscourse markers, changes involve phoric, style and code gloss markers, markers of addresser and addressee, planning and arranging, and so on. For example, metadiscourse markers of addresser and addressee are the important features along the written dimension of reader guidance, along the spoken dimension of united organization, but along the dimension of discourse presentation in the present model.

Register variation along metadiscourse dimensions

The relations among registers along Metadiscourse Dimension 1, participants’ interaction, are summarized in Figure 2 . This figure displays mean dimension scores and standard deviations of the 10 registers on Dimension 1. All spoken registers plot on the positive side, while all written registers on the negative side. Spoken registers are characterized by frequent occurrences of metadiscourse markers of addresser, addressee, saying and arguing, receiving and understanding, and defining and explaining (see Table 3 ). On the contrary, in written registers, these metadiscourse markers are of rare occurrence. In other words, speech more than writing uses metadiscourse markers to explicitly emphasize the interaction between the addresser and addressee. Most standard deviations of register scores given in Figure 2 are smaller than 1.0, and only two for unscripted speeches (1.31) and discussion broadcasts (1.61) larger than 1.0 and smaller than 2.0. This shows that along Dimension 1, texts in each register are grouped tightly around the mean score.

Results from the General Linear Models procedure reveal highly significant differences among the 10 registers along Metadiscourse Dimension 1 ( F = 167.151, p = 0.000). Tukey’s HSD post hoc tests prove that each spoken register is significantly different from each written register along Dimension 1 ( p < 0.05). Among spoken registers, there is a significant difference between each pair of registers ( p < 0.05) with the exception of unscripted speeches and casual conversations ( p = 0.189 > 0.05). Among written texts, academic discourse is significantly different from other registers ( p < 0.05). Press, general prose, and fiction share more similarities with each other ( p > 0.05). The R 2 value of 70.9 per cent shows that 70.9 per cent of the variation in the dimension scores can be accounted for by knowing the register distinctions. Metadiscourse Dimension 1 is a very powerful indicator of register differences.

Figure 3 plots the mean dimension scores of the 10 registers for Metadiscourse Dimension 2, discourse presentation. Academic prose has the highest score. General prose, scripted speeches, unscripted speeches, and public conversations have moderately high scores. These registers are more likely to use phoric markers, style markers, code gloss markers, and metadiscourse markers of introducing and analyzing, discourse, planning and arranging, and addresser and addressee to present the discourse (see Table 3 ). Press, fiction, nondiscussion broadcasts, discussion broadcasts, and casual conversations have negative scores. Metadiscourse markers of Dimension 2 seldom occur in these negative registers. All standard deviations are smaller than 1.0 except for that for academic prose which is slightly larger than 1.0 (1.47) (see Figure 3 ). For each register, most texts are grouped around the mean score tightly.

The General Linear Models procedure shows that academic prose is statistically highly different from all other registers along this dimension ( p = 0.000 < 0.001). Other positive registers are significantly differentiated from negative registers ( p < 0.05). As regards positive registers, there is no significant difference between each pair except general prose and scripted speeches ( p = 0.03 < 0.05). Similarly, negative registers share more similarities along this dimension ( p > 0.05) with the only exception of discussion broadcasts which are significantly different from press, fiction, and casual conversations ( p < 0.05). The R 2 value of 40.7 per cent indicates that there is a strong correlation between the register distinctions and the values of Metadiscourse Dimension Score 2. Metadiscourse Dimension 2 is a powerful predictor of the differences among registers.

Metadiscourse variation among registers comprises two dimensions. The relations among registers must be conceptualized with respect to both dimensions. It is inadequate and often inaccurate to compare registers in terms of only one dimension. For example, along Metadiscourse Dimension 1 (participants’ interaction), public and casual conversations are similar to each other ( p = 0.189 > 0.05). Along Metadiscourse Dimension 2 (discourse presentation), however, these two registers are highly different from each other ( p = 0.000 < 0.001). Figure 4 displays the mean dimension scores of the 10 registers on both dimensions. All registers score higher along Dimension 1 than along Dimension 2 except for academic and general prose. Besides, spoken registers have higher mean dimension scores than written registers or metadiscourse markers occur more frequently in speech than in writing. Exceptions are academic and general prose which have high scores and casual conversations which have a low score on Dimension 2. Except for these two general findings, a more detailed comparison of registers along both dimensions is not easy by manually examining the distribution of registers. Cluster analysis is introduced into the multidimensional comparison.

Mean dimension scores of the 10 registers along two metadiscourse dimensions. acad pro: academic prose; gen pro: general prose; prss: press; fict: fiction; nondis br: nondiscussion broadcasts; dis br: discussion broadcasts; scr spch: scripted speeches; unscr spch: unscripted speeches; pub conv: public conversations; cas conv: casual conversations.

Register clustering in terms of metadiscourse dimensions

In order to further explore the similarities and differences between spoken and written registers, a cluster analysis is conducted based on the mean dimension scores of the 10 registers on both metadiscourse dimensions. Registers in the same cluster display more similarities along both metadiscourse dimensions, while those of different clusters demonstrate more differences. The output of the cluster analysis is presented in Figure 5 . This dendrogram shows that the 10 registers first split into two big clusters: ‘press, fiction, general prose, nondiscussion broadcasts, academic prose’ and ‘scripted speeches, unscripted speeches, discussion broadcasts, public conversations, casual conversations.’ The first cluster is then divided into ‘academic prose’ and ‘press, fiction, general prose, nondiscussion broadcasts’ which are further grouped into ‘press and fiction’ and ‘general prose and nondiscussion broadcasts’. In the second big cluster, ‘casual conversations’ are separated from ‘scripted speeches, unscripted speeches, discussion broadcasts, public conversations’, which are divided into ‘scripted speeches and unscripted speeches’ and ‘discussion broadcasts and public conversations’.

Dendrogram from a hierarchical cluster analysis of registers in terms of metadiscourse.

Results from the cluster analysis allows different decisions about the cluster solutions. This article chooses a six-cluster solution for a detailed description of the relations among registers. Ten spoken and written registers are divided into six clusters in terms of metadiscourse: Cluster 1 (discussion broadcasts and public conversations), Cluster 2 (scripted speeches and unscripted speeches), Cluster 3 (general prose and nondiscussion broadcasts), Cluster 4 (press and fiction), Cluster 5 (casual conversations), and Cluster 6 (academic prose).

As mentioned in the ‘Introduction’ section, the register perspective of discourse analysis assumes that linguistic features are used in association with the situational context of the discourse ( Halliday and Hasan 1989 ; Biber and Conrad 2009 ). Metadiscourse markers tend to appear in a register because they are suited to the situational context of the register. Among various situational characteristics ( Biber and Conrad 2009 : 40), the use of metadiscourse across speech and writing is greatly affected by the channel, participant, and communicative purpose.

The channel or mode difference between speech and writing is one of the most basic and obvious differences in the situational context. In my cluster analysis, the first split of the 10 registers into 2 big clusters is a division between speech and writing. Scripted speeches, unscripted speeches, discussion broadcasts, public conversations, and casual conversations are spoken registers, and press, fiction, general prose, and academic prose are written registers. Taking into consideration mean dimension scores of various registers along two metadiscourse dimensions (see Figure 4 ), it can be summarized that speech and writing are significantly different in metadiscourse, and metadiscourse markers (especially those of participants’ interaction) are much more pervasive in spoken registers.

One may notice that nondiscussion broadcasts which are spoken materials in the corpus (see Table 2 ) are grouped with a written register, general prose. This is mainly because nondiscussion broadcasts in the corpus are mainly news or documents prepared or scripted beforehand ( Scannell 1991 : 1; Montgomery 2007 : 39). Discourse participants of nondiscussion broadcasts, the presenter and audience, have little control of the news or documents. In other words, nondiscussion broadcasts deviate from typical spoken registers which are part of the interaction between discourse participants ( Tannen 1983 : 80; Biber 1988 : 47; Crystal 2001 : 25–28). They are similar to written registers, or more specifically general prose, in the use of metadiscourse. In the following of this section, nondiscussion broadcasts will be included in the discussion of written registers.

The situational parameter of participants determines the cluster structure of spoken registers in terms of metadiscourse. Participants’ different degrees of shared background knowledge ( Biber and Conrad 2009 : 42) separate casual conversations from other spoken registers. In the corpus, interlocutors in casual conversations are usually friends, family members, classmates, and so on. They know each other well, and share more background knowledge than discourse participants in speeches, broadcasts, and public conversations ( Koch and Oesterreicher 1985 : 21; Biber and Conrad 2009 : 8, 37). Participants’ different degrees of interactiveness ( Biber and Conrad 2009 : 42) distinguish scripted and unscripted speeches from discussion broadcasts and public conversations. Interactiveness refers to the degree to which discourse participants interact with each other directly. In the corpus, speeches are monologues, and discussion broadcasts and public conversations are dialogues. Monologues are usually the production on the sole side of the addresser ( Enyedy and Hoadley 2006 : 418), while dialogues are constructed jointly by the addresser and addressee ( Koch and Oesterreicher 1985 : 21; Linell 2001 : 9; Arminen 2005 : 43–47). The addresser and addressee in monologues are less able to respond to each other directly than those in dialogues. Consideration of mean dimension scores of these monologues and dialogues on both metadiscourse dimensions (see Figure 4 ) reveals that dialogues use much more metadiscourse markers of participants’ interaction than monologues, and there is no significant difference between dialogues and monologues in the use of metadiscourse for discourse presentation.

Differences in communicative purposes, or more specifically informationality, divide written registers into three groups with respect to metadiscourse: academic prose; general prose and nondiscussion broadcasts; press and fiction. The major communicative purpose of informational registers, as the name suggests, is to provide information. Informational registers which are usually abstract and formal show high density and concise presentation of information. Among written registers in the corpus of this research, academic prose is the most informative, general prose and nondiscussion broadcasts are less, and press and fiction are the least ( Biber 1988 : 129–134). A further examination of mean dimension scores (see Figure 4 ) shows that more informational registers generally adopt more metadiscourse markers, especially those of discourse presentation.

(10) I mean even I understand you were talking about the children sort of being out in the dark […] (HUV) (11) The importance […] will be discussed again in a later section, after results from other attention paradigms have been summarized. (J24)

Among the 10 registers analyzed in this research, register Clusters 1 (discussion broadcasts and public conversations) and 2 (scripted and unscripted speeches) have very high mean scores on Metadiscourse Dimension 1 (participants’ interaction), and moderately high mean scores on Metadiscourse Dimension 2 (discourse presentation). For one thing, all these four registers are highly interactive dialogues (see above). For another, as institutional discourses, information and knowledge communication is an important purpose ( Arminen 2005 : 38; Heritage and Clayman 2010 : 36) of these four registers. Register Clusters 3 (general prose and nondiscussion broadcasts) and 4 (press and fiction) have markedly low scores on both dimensions. These four registers have low interactive and informational focuses (see above). Casual conversations rank among the most interactive and least informational registers ( Biber 1988 : 128), and accordingly have a markedly high mean score on the dimension of participants’ interaction and a markedly low mean score on the dimension of discourse presentation. Academic prose has a very low interaction focus and a very high information focus ( Biber 1988 : 128), and scores very low along the dimension of participants’ interaction and the highest along the dimension of discourse presentation.

This study is the first attempt at describing the overall relations among spoken and written registers in terms of metadiscourse. Using a corpus of 626 texts of 10 spoken and written registers and the method of multidimensions analysis, it identifies basic metadiscourse dimensions, and compares and clusters registers in terms of the dimensions.

The factor analysis of 271 metadiscourse markers in the corpus identifies two metadiscourse dimensions: participants’ interaction (Metadiscourse Dimension 1) and discourse presentation (Metadiscourse Dimension 2). Metadiscourse Dimension 1 is characterized by metadiscourse markers of addressee, addresser, saying and arguing, receiving and understanding, and defining and explaining. These metadiscourse markers co-occur consistently to highlight the interaction between discourse participants. Metadiscourse Dimension 2 is constituted of phoric markers, code gloss markers, style markers, and metadiscourse markers of discourse, introducing and analyzing, planning and arranging, and addresser and addressee. These frequently co-occurring metadiscourse markers mainly comment on the ongoing discourse.

Comparison of registers along metadiscourse dimensions and the results of cluster analysis show that the degrees of interactivity and informationality of registers determine their use of metadiscourse. The more interactive registers resort more to metadiscourse of participants’ interaction, and the more informational registers more to metadiscourse of discourse presentation. Ten spoken and written registers fall into six clusters in terms of metadiscourse: Cluster 1 (discussion broadcasts and public conversations), Cluster 2 (scripted speeches and unscripted speeches), Cluster 3 (general prose and nondiscussion broadcasts), Cluster 4 (press and fiction), Cluster 5 (casual conversations), and Cluster 6 (academic prose). Among them, registers of Clusters 1 and 2 are prominent along both metadiscourse dimensions as they are both interactive and informational. On the contrary, registers of Clusters 3 and 4 are prominent on neither dimension because they have low interactive and informational focuses. The highly interactive register (Cluster 5: casual conversations) is prominently marked by metadiscourse of participants’ interaction, and the highly informational register (Cluster 6: academic prose) by metadiscourse of discourse presentation.

The present study makes a contribution in regard to the theory of metadiscourse, the application of the multidimensional analysis to metadiscourse research, and the knowledge of metadiscourse use. Theoretically, it defines two dimensions of variation in metadiscourse. This two-dimensional model should continue to prove useful to the study of metadiscourse in various registers. What’s more, this model offers a classification of metadiscourse (participants’ interaction and discourse presentation) from a new perspective, that is, the co-occurrence patterns of metadiscourse markers in discourse. Methodologically, this research proves that the multidimensional analysis is a powerful method for the study of register variation in metadiscourse. In addition to extracting metadiscourse dimensions using factor analysis, it incorporates cluster analysis in the multidimensional analysis to cluster registers in terms of both metadiscourse dimensions. This resolves the difficulties in multidimensional description of register variation caused by distribution disparities of registers along different dimensions. Furthermore, this study suggests that metadiscourse can be added into features or dimensions distinguishing English registers recognized by Biber (1988) . As for the use of metadiscourse, the present research analyzes and compares metadiscourse across 10 spoken and written registers, which provides a fuller description of English metadiscourse.

Despite its contribution, it must be emphasized that the findings about register variation in metadiscourse represent the prototypical or typical use of metadiscourse in the 10 broad registers. In each register, there is a continuous range of variation among sub-registers. It is advisable to further explore metadiscourse variation among sub-registers in the future. Besides, this study which aims at a comprehensive investigation of metadiscourse can be extended by using a larger corpus, examining more registers, and analyzing more metadiscourse markers. In this research, most texts are extracts instead of full texts, digital registers are not considered, and only part of the corpus was read manually to find potential metadiscourse markers. The extensions, however, may not be easily achieved because identifying metadiscursive expressions in texts which is the basis of almost all metadiscourse research can only be done manually. Identifying metadiscursive expressions in a large corpus can be very time consuming and labor intensive. To study metadiscourse more comprehensively and understand the use of metadiscourse more fully, the development of more efficient methods of metadiscourse identification are worthy of future research.

For example, Ädel (2010) carried out a study comparing the use of personal metadiscourse in spoken lectures and written essays.

Examples of this article are from its corpus (see ‘Corpus’ section). Text numbers at the end of the examples are from the original corpora (see ‘Corpus’ section).

BNC is not used for the written materials mainly because we had finished the metadiscourse analysis in written registers using F-LOB as the corpus before I looked for spoken materials for the metadiscourse analysis in spoken registers.

In F-LOB, this register is named ‘learned writing’. This article uses ‘academic prose’ because it is a more familiar label now.

Biber (1988 : 170) uses ‘text types’ to refer to ‘groupings of texts that are similar in their linguistic form’.

Supplementary material is available at Applied Linguistics online.

I am deeply grateful to the analysts for doing part of the corpus analysis, and anonymous reviewers and editors for their many insightful comments on previous versions of this article.

Conflict of interest statement . None declared.

Ädel A. 2006 . Metadiscourse in L1 and L2 English . John Benjamins Publishing Company .

Google Scholar

Google Preview

Ädel A. 2010 . ‘ Just to give you kind of a map of where we are going : A taxonomy of metadiscourse in spoken and written academic English ,’ Nordic Journal of English Studies 9 : 69 – 97 .

Ädel A. 2017 . ‘ Remember that your reader cannot read your mind: Problem/solution–oriented metadiscourse in teacher feedback on student writing ,’ English for Specific Purposes 45 : 54 – 68 .

Aguilar M. 2008 . Metadiscourse in Academic Speech: A Relevance–Theoretic Approach . Peter Lang .

Anthony L. 2013 . ‘AntConc (Version 3.2.4w),’ available at www.laurenceanthony.net Accessed 16 March 2013.

Arminen I. 2005 . Institutional Interaction: Studies of Talk at Work . Ashgate .

Biber D. 1988 . Variations across Speech and Writing . Cambridge University Press .

Biber D. 1995 . Dimensions of Register Variation: A Cross-Linguistic Comparison . Cambridge University Press .

Biber D , Conrad S. 2009 . Register, Genre, and Style . Cambridge University Press .

Biber D. , Johansson S. , Leech G. , Conrad S. , Finegan E. . 1999 . Longman Grammar of Spoken and Written English . Longman .

Bouziri B. 2021 . ‘ A tripartite interpersonal model for investigating metadiscourse in academic lectures ,’ Applied Linguistics 42 : 970 – 89 .

Costello A. B , Osborne J. W. . 2005 . ‘ Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis,’ Practical Assessment , Research & Evaluation 10 : 1 – 9 .

Crismore A. 1989 . Talking with Readers: Metadiscourse as Rhetorical Act . Peter Lang .

Crismore A. V. O. N. , Markkanen R. , Steffensen M. S. . 1993 . ‘ Metadiscourse in persuasive writing: A study of texts written by American and Finnish university students ,’ Written Communication 10 : 39 – 71 .

Crystal D. 2001 . Language and the Internet . Cambridge University Press .

Enyedy N , Hoadley M. C. . 2006 . ‘ From dialogue to monologue and back: Middle spaces in computer-mediated learning ,’ International Journal of Computer-Supported Collaborative Learning 1 : 413 – 39 .

Everitt B. S. , Landau S. , Leese M. , Stahl D. . 2011 . Cluster Analysis , 5th edn. John Wiley & Sons, Ltd .

Halliday M. A. K. , Hasan R. . 1989 . Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective . Oxford University Press .

Harris Z. 1959 . ‘ The transformational model of language structure ,’ Anthropological Linguistics 1 : 27 – 9 .

Heritage J , Clayman S. . 2010 . Talk in Action: Interactions, Identities, and Institutions . Wiley-Blackwell .

Hyland K. 1998 . ‘ Exploring corporate rhetoric: Metadiscourse in the CEO’s letter ,’ Journal of Business Communication 35 : 224 – 44 .

Hyland K. 2005 . Metadiscourse: Exploring Interaction in Writing . Continuum .

Hyland K. 2017 . ‘ Metadiscourse: What is it and where is it going? ,’ Journal of Pragmatics 113 : 16 – 29 .

Jiang F , Hyland K. . 2016 . ‘ Nouns and academic interactions: A neglected feature of metadiscourse ,’ Applied Linguistics 2016 : 1 – 25 .

Khabbazi-Oskouei L. 2013 . ‘ Propositional or non-propositional, that is the question: A new approach to analyzing “interpersonal metadiscourse” in editorials ,’ Journal of Pragmatics 47 : 93 – 107 .

Koch P , Oesterreicher W. . 1985 . ‘ Sprache der Nähe–Sprache der Distanz. Mündlichkeit und Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgeschichte ,’ Romanistisches Jahrbuch 36 : 15 – 43 .

Lee D. 2001 . ‘ Genres, registers, text types, domains, and styles: Clarifying the concepts and navigating a path through the BNC jungle ,’ Language Learning and Technology 3 : 37 – 72 .

Linell P. 2001 . Approaching Dialogue: Talk, Interaction and Contexts in Dialogical Perspectives . John Benjamins Publishing Company .

Lyons J. 1977 . Semantics (vol. 1 ). Cambridge University Press .

Mauranen A. 1993a . ‘ Contrastive ESP rhetoric: Metatext in Finnish–English economics texts ,’ English for Specific Purposes 12 : 3 – 22 .

Mauranen A. 1993b . Cultural Differences in Academic Rhetoric: A Textlinguistic Study . Peter Lang .

Mauranen A. 2010 . ‘ Discourse reflexivity—A discourse universal? The case of ELF ,’ Nordic Journal of English Studies 9 : 13 – 40 .

Moisl H. 2015 . Cluster Analysis for Corpus Linguistics . De Gruyter Mouton .

Montgomery M. 2007 . The Discourse of Broadcast News . Routledge .

Scannell P. 1991 . Broadcast Talk . Sage .

Sinclair J. 2005 . ‘Language as a string of beads: Discourse and the M-word’ in Tognini-Bonelli E. and Camiciotti G. D. L. (eds): Strategies in Academic Discourse . John Benjamins , pp. 163 – 8 .

Tannen D. 1983 . ‘Oral and literate strategies in spoken and written discourse’ in Bailey R. W. and Fosheim R. M. (eds): Literacy for Life: The Demand for Reading and Writing . The Modern Language Association , pp. 79 – 96 .

Tse P , Hyland K. 2008 . ‘“ Robot Kung Fu”: Gender and professional identity in biology and philosophy reviews ,’ Journal of Pragmatics 40 : 1232 – 48 .

Vande Kopple W. J. V. 1985 . ‘ Some exploratory discourse on metadiscourse ,’ College Composition and Communication 36 : 82 – 93 .

Williams J. M. 1981 . Style: Ten Lessons in Clarity and Grace . Pearson .

Zhang M. 2016 . ‘ A multidimensional analysis of metadiscourse markers across written registers ,’ Discourse Studies 18 : 204 – 22 .

Zhang M. , Sun W. , Peng H. , Gan Q. , Yu B. . 2017 . ‘ A multidimensional analysis of metadiscourse markers across spoken registers ,’ Journal of Pragmatics 117 : 106 – 18 .

Man Zhang is a Professor of Linguistics at the School of Foreign Languages, Central China Normal University, Wuhan, P. R. China, and a doctoral candidate at the Faculty of Language, Literature and Humanities, Humboldt University of Berlin, Berlin, Germany. Her research interests include discourse analysis, corpus linguistics, computational linguistics, functional linguistics, and language teaching. Her recent publications on metadiscourse appeared in Discourse Studies , Journal of Pragmatics , Journal of Quantitative Linguistics , and so on.

Supplementary data

Email alerts, citing articles via, looking for your next opportunity.

• Recommend to your Library

Affiliations

• Online ISSN 1477-450X
• Print ISSN 0142-6001
• Copyright © 2024 Oxford University Press
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Speech in Linguistics

• An Introduction to Punctuation
• Ph.D., Rhetoric and English, University of Georgia
• M.A., Modern English and American Literature, University of Leicester
• B.A., English, State University of New York

In linguistics , speech is a system of  communication  that uses spoken words  (or sound symbols ). 

The study of speech sounds (or spoken language ) is the branch of linguistics known as phonetics . The study of sound changes in a language is phonology . For a discussion of speeches in rhetoric and oratory , see Speech (Rhetoric) .

Etymology:  From the Old English, "to speak"

Studying Language Without Making Judgements

• "Many people believe that written language is more prestigious than spoken language--its form is likely to be closer to Standard English , it dominates education and is used as the language of public administration. In linguistic terms, however, neither speech nor writing can be seen as superior. Linguists are more interested in observing and describing all forms of language in use than in making social and cultural judgements with no linguistic basis." (Sara Thorne, Mastering Advanced English Language , 2nd ed. Palgrave Macmillan, 2008)

Speech Sounds and Duality

• "The very simplest element of speech --and by 'speech' we shall henceforth mean the auditory system of speech symbolism, the flow of spoken words--is the individual sound, though, . . . the sound is not itself a simple structure but the resultant of a series of independent, yet closely correlated, adjustments in the organs of speech." ( Edward Sapir , Language: An Introduction to the Study of Speech , 1921)
• "Human language is organized at two levels or layers simultaneously. This property is called duality (or 'double articulation'). In speech production, we have a physical level at which we can produce individual sounds, like n , b and i . As individual sounds, none of these discrete forms has any intrinsic meaning . In a particular combination such as bin , we have another level producing a meaning that is different from the meaning of the combination in nib . So, at one level, we have distinct sounds, and, at another level, we have distinct meanings. This duality of levels is, in fact, one of the most economical features of human language because, with a limited set of discrete sounds, we are capable of producing a very large number of sound combinations (e.g. words) which are distinct in meaning." (George Yule, The Study of Language , 3rd ed. Cambridge University Press, 2006)

Approaches to Speech

• "Once we decide to begin an analysis of speech , we can approach it on various levels. At one level, speech is a matter of anatomy and physiology: we can study organs such as tongue and larynx in the production of speech. Taking another perspective, we can focus on the speech sounds produced by these organs--the units that we commonly try to identify by letters , such as a 'b-sound' or an 'm-sound.' But speech is also transmitted as sound waves, which means that we can also investigate the properties of the sound waves themselves. Taking yet another approach, the term 'sounds' is a reminder that speech is intended to be heard or perceived and that it is therefore possible to focus on the way in which a listener analyzes or processes a sound wave." (J. E. Clark and C. Yallop, An Introduction to Phonetics and Phonology . Wiley-Blackwell, 1995)

Parallel Transmission

• "Because so much of our lives in a literate society has been spent dealing with speech recorded as letters and text in which spaces do separate letters and words, it can be extremely difficult to understand that spoken language simply does not have this characteristic. . . . [A]lthough we write, perceive, and (to a degree) cognitively process speech linearly--one sound followed by another--the actual sensory signal our ear encounters is not composed of discretely separated bits. This is an amazing aspect of our linguistic abilities, but on further thought one can see that it is a very useful one. The fact that speech can encode and transmit information about multiple linguistic events in parallel means that the speech signal is a very efficient and optimized way of encoding and sending information between individuals. This property of speech has been called parallel transmission ." (Dani Byrd and Toben H. Mintz, Discovering Speech, Words, and Mind . Wiley-Blackwell, 2010)

Oliver Goldsmith on the True Nature of Speech

• "It is usually said by grammarians , that the use of language is to express our wants and desires; but men who know the world hold, and I think with some show of reason, that he who best knows how to keep his necessities private is the most likely person to have them redressed; and that the true use of speech is not so much to express our wants, as to conceal them." (Oliver Goldsmith, "On the Use of Language." The Bee , October 20, 1759)

Pronunciation: SPEECH

• Duality of Patterning in Language
• Phonology: Definition and Observations
• Definition and Examples of Productivity in Language
• What Is Phonetics?
• Phonological Segments
• Spoken English
• Definition of Voice in Phonetics and Phonology
• What Is a Phoneme?
• What Are Utterances in English (Speech)?
• Sound Symbolism in English: Definition and Examples
• Definition and Examples of Linguists
• What Is Graphemics? Definition and Examples
• Assimilation in Speech
• Phoneme vs. Minimal Pair in English Phonetics
• Suprasegmental Definition and Examples
• An Introduction to Semantics

Relationship and Difference Between Speech and Writing in Linguistics

Back to: Pedagogy of English- Unit 4

Many differences exist between the written language and the spoken language. These differences impact subtitling which is a practice that has become highly prevalent in the modern age. It is a process used to translate what the speaker is saying for those of other languages or who are deaf.

The main difference between written and spoken languages is that written language is comparatively more formal and complex than spoken language. Some other differences between the two are as follows:

Writing is more permanent than the spoken word and is changed less easily. Once something is printed, or published on the internet, it is out there for the world to see permanently. In terms of speaking, this permanency is present only if the speaker is recorded but they can restate their position.

Apart from formal speeches, spoken language needs to be produced instantly. Due to this, the spoken word often includes repetitions, interruptions, and incomplete sentences. As a result, writing is more polished.

Punctuation

Written language is more complex than spoken language and requires punctuation. Punctuation has no equivalent in spoken language.

Speakers can receive immediate feedback and can clarify or answer questions as needed but writers can’t receive immediate feedback to know whether their message is understood or not apart from text messages, computer chats, or similar technology.

Writing is used to communicate across time and space for as long as the medium exists and that particular language is understood whereas speech is more immediate.

Use of Slang

Written and spoken communication uses different types of language. For instance, slang and tags are more often used when speaking rather than writing.

Speaking and listening skills are more prevalent in spoken language whereas writing and reading skills are more prevalent in written language.

Tone and pitch are often used in spoken language to improve understanding whereas, in written language, only layout and punctuation are used.

These are the major differences between spoken language and written language.

Hate speech and disinformation in South Africa’s elections: big tech make it tough to monitor social media

Professor Emeritus, Rhodes University, Rhodes University

Disclosure statement

Guy Berger has received funding from the thinktank Research ICT Africa, where he is a Distinguished Research Fellow.

Rhodes University provides funding as a partner of The Conversation AFRICA.

View all partners

There’s a growing global movement to ensure that researchers can get access to the huge quantity of data assembled and exploited by digital operators.

Momentum is mounting because it’s becoming increasingly evident that data is power. And access to it is the key – for a host of reasons, not least transparency, human rights and electoral integrity.

But there’s currently a massive international asymmetry in access to data.

In the European Union and the US, some progress has been made. For example, EU researchers studying risks have a legal right of access. In the US too, some companies have taken voluntary steps to improve access.

The situation is generally very different in the global south.

The value of data access can be seen vividly in the monitoring of social media during elections. South Africa is a case in point. A powerful “big data” analysis was recently published about online attacks on women journalists there, raising the alarm about escalation around – and after – the election on 29 May.

A number of groups working with data are attempting to monitor hate speech and disinformation on social media ahead of South Africa’s national and provincial polls. At a recent workshop involving 10 of these initiatives, participants described trying to detect co-ordinated “information operations” that could harm the election, including via foreign interference.

But these researchers can’t get all the data they need because the tech companies don’t give them access.

This has been a concern of mine since I first commissioned a handbook about harmful online content – Journalism, Fake News & Disinformation: Handbook for Journalism Education and Training – six years ago. My experience since then includes overseeing a major UN study called Balancing Act: Countering Digital Disinformation While Respecting Freedom of Expression .

Over the years, I’ve learnt that to dig into online disinformation, you need to get right inside the social media engines. Without comprehensive access to the data they hold, you’re left in relative darkness about the workings of manipulators, the role of misled punters and the fuel provided by mysterious corporate algorithms.

Looking at social media in the South African elections, the researchers at the recent workshop shared how they were doing their best with what limited data they had. They were all monitoring text on social platforms. Some were monitoring audio, while a few were looking at “synthetic content” such as material produced with generative AI.

About half of ten initiatives were tracking followers, impressions and engagement. Nearly all were checking content on Twitter; at least four were monitoring Facebook; three covered YouTube; and two included TikTok.

WhatsApp was getting scant attention. Though most messaging on the service is encrypted, the company knows (but doesn’t disclose) which registered user is bulk sending content to which others, who forwards this on, whether group admins are active or not, and a host of other “metadata” details that could help monitors to track dangerous trajectories.

But the researchers can’t do the necessary deep data dives. They’ve set out the difficult data conditions they work under in a public statement explaining how they are severely constrained in their access to data.

One data source they use is expensive (and limited) packages from marketing brokers (who in turn have purchased data assets wholesale from the platforms).

A second source is from analysing published posts online (which excludes in-group and WhatsApp communications). Using scraped data is limited and labour-intensive. Findings are superficial. And it’s risky: scraping is forbidden in most platforms’ terms of use.

None of the researchers covering South Africa’s elections have direct access to the platforms’ own Application Programme Interfaces (APIs). These gateways provide a direct pipeline into the computer servers hosting data. This major resource is what companies use to profile users, amplify content, target ads and automate content moderation. It’s an essential input for monitoring online electoral harms.

In the EU, the Digital Services Act enables vetted researchers to legally demand and receive free, and potentially wide-ranging, API access to search for “systemic risks” on the platforms.

It’s also more open in the US. There, Meta, the multinational technology giant that owns and operates Facebook, Instagram and WhatsApp, cherrypicked 16 researchers in the 2020 elections (of which only five projects have published their findings). The company has subsequently outsourced the judging of Facebook and Instagram access requests (from anywhere worldwide) to the University of Michigan .

One of the South African researchers tried that channel, without success.

Other platforms such as TikTok are still making unilateral decisions, even in the US, as to who has data access.

Outside the EU and the US, it’s hard even to get a dialogue going with the platforms.

The fightback

Last November, I invited the bigger tech players to join a workshop in Cape Town on data access and elections in Africa. There was effectively no response .

The same pattern is evident in an initiative earlier this year by the South African National Editors’ Forum. The forum suggested a dialogue around a human rights impact assessment of online risks to the South African elections. They were ignored .

Against this background, two South African NGOs – the Legal Resources Centre and the Campaign for Free Expression – are using South Africa’s expansive Promotion of Access to Information Act to compel platforms to disclose their election plans.

But the companies have refused to respond, claiming that they do not fall under South African jurisdiction. This has led to appeals being launched to the country’s Information Regulator to compel disclosures.

Further momentum for change may also come from Unesco, which is promoting international Guidelines for the Governance of Digital Platforms. These highlight transparency and the issue of research access. Unesco has also published a report that I researched titled Data Sharing to Foster Information as a Public Good.

In the works is an incipient African Alliance for Access to Data , now involving five pan-African formations. This coalition (I’m interim convenor) is engaging the African Union on the issues.

But there’s no guarantee yet that all this will lead the platforms to open up data to Africans and researchers in the global south.

• Social media
• South Africa
• Hate speech
• Disinformation
• Online attacks
• South Africa election 2024

Data Manager

Research Support Officer

Director, Social Policy

Head, School of Psychology

Senior Research Fellow - Women's Health Services