Hypotheses about the spoken language-written language relation

Compared to spoken language, written language has been traditionally neglected as an object of scientific enquiry. Spanning the time from Aristotle to de Saussure evolutionary, philosophical and linguistic arguments have been put forward to underpin the alleged primacy of the spoken language. These ideas have affected deaf education throughout the centuries (cf. Plaza-Pust 2016). More recently, however, alternative proposals to the traditional derivational view of the speech- print relation have been put forward, so that, today, three positions can be distinguished in the literature concerning the relation between the written language and the spoken language (Durscheid 2006: 35), namely (a) the Dependence hypothesis, (b) the Autonomy hypothesis, and (c) the Interdependence hypothesis. We will briefly summarise the main arguments of each.

Dependence Hypothesis. The primacy of the spoken language over reading and writing lies at the core of the so-called Dependence hypothesis, which corresponds with the traditional derivational view of print. The generalised attribution of a primary status to speech is also associated with the idea that the spoken modality of expression reflects thought directly whereas print would do so only indirectly (Durscheid 2006: 14 mentions Rousseaus’s view and also Hermann Paul’s that writing would not be adequate to language). In support of the Dependence hypothesis, linguistic, evolutionary, logical, philosophical and functional arguments have been put forward (Durscheid 2006: 36).

Advocates of the Dependence hypothesis tend to emphasise that print serves only a limited range of functions (Lyons 1987 pace Durscheid 2006: 37). While scripts are regarded as a means to visualise speech, their limitations are acknowledged as not all characteristics of the spoken language are represented in script (suprasegmental elements, for example, cannot be fully expressed). Hence the view that the object of linguistic study is not the written but the spoken form, advocated by De Saussure (1972), and the assumption about their unequal status, whereby speech is attributed a primary and writing a secondary status (for example, in Daniels 1996: 1; cf. Primus & Neef 2004: 133 for a discussion).

As for sound-letter correspondences determining the relation of alphabetic writing systems and spoken languages, proponents of the Dependence hypothesis maintain that the units of the writing system (graphemes) are defined in relation to the units identified for speech (phonemes). Because the sound-letter relation is conceived of in a unidirectional manner, constraints at the graphematic level remain unaccounted for as are letter-to-sound correspondences and their relevance in the reading process (Primus & Neef 2004: 132).

Turning to the evolutionary argument, advocates of the Dependence hypothesis emphasise that writing qua socio-genetic phenomenon is not a natural phenomenon, but must be developed as a cultural product. It is argued further that learning to read and write are not spontaneous processes but are bound to formal instruction, which typically takes place after the acquisition of the spoken language. The secondary status attributed to writing systems is also commonly underpinned by the logical argument that spoken languages exist without written languages, whereas the latter would not exist without the former. Following this line of argumentation the very possibility of attaining the written language without having learned the related spoken language before is not conceived of.

Autonomy hypothesis. For advocates of the Autonomy hypothesis the written language is an object of scientific enquiry in its own right. Implicit to this claim is the assumption that the written language and the spoken language have equal status. In contrast to the derivational view, proponents of the Autonomy hypothesis highlight the natural character of the processes that shape the development of writing systems. Primus and Neef (2004: 134), for example, maintain that these processes are “comparable to those which shape language itself ... As a consequence, mature writing systems and alphabets, which have been used for a long time by large communities for encoding a specific language, are as natural as any spoken language.” Further, autonomy is also advocated regarding the acquisition of the written language without prior access to the spoken language. It is interesting to note that deaf children’s acquisition of the written language is often mentioned as evidence that would provide support for this assumption. With respect to production and comprehension, advocates of the Autonomy hypothesis argue that (skilled) reading and writing processes are not mediated by spoken language. In its strictest version, the Autonomy hypothesis maintains that there is no relation between the spoken and the written form (Neef & Primus 2001: 353). However, as is explained next, less radical approaches tend to emphasise the equal status of the two languages rather than a lack of a connection between the two systems. This position is referred in the literature as the Interdependence hypothesis (not to be confounded with Cummins’ Interdependence hypothesis discussed in section 1.3.2.2).

Interdependence hypothesis. Between the two opposed views of autonomy vs. dependence a third approach, the Interdependence hypothesis, is advocated by those who argue that although the written language is an object of scientific enquiry in its own right, correspondences between the two systems must be acknowledged. Gunther (2003: 39), for example uses the notion of “relative autonomy” to emphasise, on the one hand, the equal status of spoken, signed and written languages that can be acquired independently from each other. On the other hand, the author also remarks on the structural relationship between the written language and the spoken language.

Proponents of the Interdependence hypothesis not only attribute an equal status to spoken language and written language, they also assume that the nature of the relation is reciprocal rather than uni-directional or non-existent. This view implies that correspondence rules regulating the relation between the spoken and written system comprise regularities of the writing system that might be phonologically or graphematically based or biunique (that is, one-to-one in both directions) (Neef & Primus 2001: 365). Implicit to this view is the assumption that some constraints of writing systems have no correspondence in the spoken language system. Indeed, graphematic features are assumed to be part of lexical entries. In other words, it is assumed that the lexicon contains a graphematic component.

In recognising the mutual influence between the spoken language and the written language as well as their respective specific characteristics the Interdependence hypothesis not only fits well with their historical development as the two developed independently, although in connection with each other. It also fits well with the dual model of reading and writing processing (Jimenez-Gonzalez & Muneton-Ayala 2002: 42f.; Sprenger-Charolles & Bechennec 2004: 14). According to this model, two main processes or routes are involved in word writing (and word reading), namely, (a) the phonological route (a word is written based on its phonological form through the conversion of phonemes to graphemes), and (b) an orthographic or graphematic route (a word is written based on the orthographic form selected from the orthographic or graphematic lexicon). Both routes have been found to be used also in the production of Spanish or Italian, among the most orthographically transparent languages, with only few irregularities in phoneme-grapheme correspondences (Sprenger-Charolles & Bechennec 2004: 14).

Turning to acquisition, the inter-dependence model also opens a new perspective in the controversy about the status of the written language in deaf individuals’ language acquisition. In this framework, learners are expected to profit from knowledge about the correspondence rules that constrain the relation of the written language to the spoken language. What is more, the equal status attributed to the two languages and their autonomy qua systems in their own right also allows for the conception of alternative routes in their acquisition: not only with respect to which system is acquired first (for example, learners might start out by learning a specific writing system, attaining the related spoken language only later) but also concerning preferred processing routes (phonemic, graphemic or both). It is this potential of using alternative routes that is particularly interesting in the case of deaf children whose acquisition of the spoken language is bound to be delayed if not truncated owing to their hearing loss.