Cognitive versus generative models

As we have seen, the formal approach assumes the modularity thesis and, within this, the autonomy of syntax thesis. Cognitive approaches, on the other hand, assume that language is not an encapsulated system but a system embedded within and inextricable from generalised cognitive processes.

Lexicon-grammar continuum versus autonomous syntax

 It follows from this view that, in the cognitive model, syntax is not autonomous. Instead, the syntax forms part of a continuum together with lexicon and morphology. This continuum consists of symbolic units of varying shapes and sizes. In the formal model, the syntactic component mediates between form and meaning, whereas in the cognitive model, grammatical structures are just another kind of meaningful symbolic unit, albeit of a schematic nature. Furthermore, in formal approaches, particularly Chomsky’s approach, the symbols that operate within the syntax or ‘Computational System’ are meaningless. Indeed, many of the features that drive the Merge and Move operations in the Minimalist Program are described as ‘uninterpretable’ features, which have to be eliminated in the course of the derivation to avoid an ungrammatical output. Category selection features work in this way, for example: a verb ‘arrives’ in the syntax with selection features (e.g. the verb kiss selects a complement with the category N), and this selection feature is eliminated from the representation when the verb is merged with a complement bearing the appropriate feature. If this feature is not eliminated (in other words, if the verb fails to get its NP object), the result is ungrammatical. These features thus serve only to create a well-formed grammatical output, regardless of the semantics of that output, and the syntax operates blindly and automatically over these features. As we have seen, in cognitive approaches the symbolic units that comprise the grammatical system are meaningful and serve a structuring function.

Inventory versus derivational system

Despite the fact that cognitive approaches to grammar share with formal approaches the ultimate objective of modelling speaker knowledge, and despite the fact that both assume a ‘dynamic’ model (that is, not a static body of know ledge), the architecture of the two models differs considerably in nature. While the formal model posits a computational system that generates (builds or derives) well-formed grammatical structures without recourse to meaning, the cognitive model posits an inventory of symbolic units containing ‘schematic templates’. These templates are formed as a consequence of regular use and are thus entrenched. When a speaker forms or interprets new structures, he or she does so not by applying a set of rules or principles, but by comparing the new structure with existing templates, and by taking into account the goals of the communicative exchange, the context and so on. In this sense, the cognitive model is a problem-solving model. While formal models capture generalisations and define well-formedness (or grammaticality) as the output of precisely stated rules and principles, the cognitive model captures generalisations and defines well-formedness (or conventionality) as the result of a categorisation process.

The status of constructions

 As we have seen, the term construction is used rather differently in cognitive and formal approaches. In the cognitive approach, it refers to a symbolic unit, which may be as small as a word or as big as a clause, that is stored ‘whole’ within the inventory of symbolic units that represents the speaker’s knowledge of language. In the cognitive approach, the construction is primitive, in the sense that it does not represent the output of any more fundamental linguistic unit or process. In formal approaches, the term ‘construction’ is usually applied only to clauses, and, in derivational theory, carries with it the sense that the structure has been ‘built’ by the application of grammatical structure building rules and transformational rules. In this type of model, the construction is epiphenomenal, because it emerges as the output of more fundamental primitives and processes (the ‘words and rules’ model). As we saw in Chapter 21, however, Kay and Fillmore’s Construction Grammar is a ‘broadly generative’ theory that takes a rather different view of constructions in the sense that constructions, although inheriting certain properties from other constructions, are ‘stored whole’ rather than built from syntactic rules. In this respect, the Construction Grammar view of constructions shares more in common with the cognitive model than with other generative models.

Schemas versus rules

A further point of contrast between the two theories concerns the distinction between schemas and rules, which follows from a number of points that we have already discussed. Consider the Cognitive Grammar schema for plural nouns in (3).

The question that arises here concerns how the presence of a schema like (3) in the grammar is different from a derivational rule, since both aim to capture the same aspect of speaker knowledge. The difference lies in the directionality of the relationship between the schema or rule on the one hand, and the specific expressions that correspond to it on the other. In the generative model, the rule precedes and thus determines the specific expressions that instantiate it. In the cognitive model, the schema does not give rise to the instance but follows from it: the schema represents a pattern that emerges from entrenched units as a consequence of usage. Of course, novel uses represent an exception to this generalisation in the sense that they are sanctioned by existing schemas. In these ways, both models account for well-formedness.

Redundancy versus economy

 In the cognitive model, generalisations result from recurring patterns of usage that enable the speaker to arrive at a ‘higher-order’ schema. This means that both schemas (the cognitive counterpart of rules) and instances of those schemas (lists of specific constructions) coexist in the grammar, and the schema is therefore an expression of the generalisation that emerges from pat terns of usage. In contrast, generative linguists argue that that forms that can be derived from the application of a generalised rule need not be listed in the grammar. For example, if the rule N s derives plural nouns, then specific instances like philanderers, lovers and deceivers need not be listed in the grammar in addition to their singular counterparts, because the singular nouns plus the generalised rule can straightforwardly derive the plural forms. This rule/list dichotomy is motivated on the basis of economy: it is argued that language must be a maximally economical system in order to be acquired and manipulated so rapidly, hence the model should avoid redundancy. Indeed, this economy-driven approach lies at the core of Chomsky’s Minimalist approach.

Conventionality versus regularity

 A related difference between the formal and cognitive approaches concerns the nature of the phenomena each model attempts to account for. Formal approaches to grammar have tended to focus on the statement of general rules that account for grammaticality or well-formedness in any given language, and in human language in general. For this reason, generative theories of grammar tend not to be concerned with ‘conventional’ or ‘fixed’ expressions, just as formal theories of meaning have not been concerned with ‘non-compositional’ or ‘figurative’ language. Since conventional or idiomatic expressions like by and large or kick the bucket clearly have complex syntactic structure, they are atypical lexical items. Since such expressions often fail to conform to general patterns of syntactic structure, they are not accounted for by this component of the grammar either. In the formal approach, such expressions are considered peripheral and uninteresting because they do not reveal general and productive patterns. Instead, the formal model focuses upon ‘core’ phenomena (word order, major clause types, case and agreement patterns and so on). This is because generalisation is a primary objective of this approach, which emerges as a consequence of its central research goal, which is to characterise Universal Grammar. In contrast, the cognitive approach views conventional and idiomatic expressions as a central part of what it means to know and to use a language. Indeed, these ‘irregular’ expressions are not viewed as unusual or problematic because the cognitive model does not assume a rule/list dichotomy. Instead, all expressions, ‘regular’ or ‘irregular’, form part of a speaker’s inventory of linguistic knowledge and must be accounted for.

‘Scaffolding’ versus ‘building blocks’

This point of contrast relates to the status of compositional structure within the model. As we have seen, the generative model assumes that rules give rise to constructions, which Langacker (1987) describes in terms of the building block metaphor. In other words, formal models view linguistic elements as having a componential structure: elements from speech sounds to sentences are viewed as having a complex internal structure, which may consist of structural ‘building blocks’ like articulatory features, morphemes or grammatical categories, or which may consist of semantic ‘building blocks’ like semantic primitives or meaning components. In Langacker’s view, while these ‘building blocks’ may serve a useful practical function as classificatory features, and may even have cognitive reality, they are epiphenomenal. In other words, they are a ‘symptom’ of the status of that linguistic expression within a complex network of meanings and forms, but are not themselves the foundations of either meaning or structure within linguistic expressions.

In contrast, as we have seen, the usage-based model holds that entrenched instances give rise to schemas. Despite this important difference, Langacker’s model of grammatical constructions acknowledges that complex structures are recognised by speakers as having compositional structure. Indeed, it is the recognition of recurring structural patterns that enables speakers to create novel grammatical constructions. Langacker proposes an alternative to the building-block metaphor that encompasses both compositional and non-com positional units: the scaffolding metaphor. In Cognitive Grammar, component structures are described as immanent in the complex grammatical construction, regardless of whether the compositionality is recognised by the speaker. Langacker argues that entrenchment decreases the salience of compositionality. For example, we are less aware of the well-entrenched noun computer as a complex construction than we are of a less well-entrenched or novel instance like striver. The compositional structure of a grammatical construction may be essential to the initial creation or construction of that expression, but once the construction is entrenched and gains the status of a unit, this com positional scaffolding is no longer required. Despite this, the compositional structure remains immanent: we may still recognise the compositionality of well-entrenched units, but it does not follow that we ‘build them from scratch’ each time we use them. The fact that certain complex constructions do not conform to the prototypical patterns of compositionality does not present a problem in this model. For example, we might argue that the compound noun bluebottle or the idiomatic expression have a butchers at represent cases where the individual components are no longer recognised as making a contribution to the construction as a whole, and that these expressions have therefore been reanalysed as simplex units, at least at the semantic pole.

Constraints on the model

Cognitive and generative approaches also differ to a considerable extent in terms of the constraints placed upon the model. Because of its emphasis on economy and generalisation, the formal model places strict constraints upon grammatical constructions and processes. This is particularly evident in its emphasis on the relatedness of constructions. For example, the transformational model assumes that all clause types are constructed according to the same general principles and share a similar underlying structure. Furthermore, it is assumed that non-canonical clause types like interrogative clauses, passive clauses and cleft clauses are related to, and therefore derived from, more basic underlying clause structures. In order to preserve these assumptions, the trans formational model admits ‘invisible’ and semantically empty elements. Invisible elements lack phonetic realisation but are thought to be present for semantic or structural reasons. Consider the examples in (4).

Example (4a) contains an embedded clause, and the NP Lily is the subject of the embedded clause (she is doing the seeing). In the transformational model, example (4b) is also thought to include an embedded subject (interpreted as co referential with George) that has no phonetic realisation. This invisible embedded subject is represented by the underscore. This assumption preserves the view that both examples share a parallel structure. Semantically ‘empty’ elements include so-called ‘dummy’ elements. For example, the ‘dummy’ subject it in it surprised her that he turned up at all has no referential content. We also saw in Chapter 14 that the auxiliary verb do is described as a ‘dummy’ auxiliary since it is conditioned by certain grammatical requirements but does not bring its own contribution to the clause in terms of aspect or voice. In sum, while the formal model places severe constraints on grammatical constructions and processes, it allows a proliferation of ‘invisible’ and ‘dummy’ elements in order to preserve generalisations.

In contrast, the cognitive model adopts the opposite position: ‘invisible’ or ‘semantically empty’ elements are not permitted, but constructions, related and unrelated, proliferate. For example, in Cognitive Grammar, the Content Requirement prohibits invisible or semantically empty elements, although symbolic units can be implicit (for example, class schemas). However, even implicit symbolic units are meaningful, albeit schematic. The Content Requirement also prohibits abstract ‘underlying’ structures. However, the cognitive model does not emphasise constraints upon grammatical constructions, which proliferate. Because the cognitive model views redundancy as natural and is less concerned with generalisation, it requires less theoretical machinery.

Emphasis on formalism

This brings us to our final point of contrast in this section. An important difference between formal and cognitive approaches, as the term ‘formal approach’ itself suggests, is a different degree of emphasis on formalism. Formalism in linguistics is the practice of adopting a metalanguage for the description of natural language phenomena, and often involves the manipulation of abstract symbols and rules. As we saw in Chapter 13, formal semantics adopts logic as a metalanguage for the description of linguistic meaning, and the tree diagrams, transformational rules and abstract features of Chomsky’s approach to syntactic theory are also components of a formal metalanguage for describing the grammatical properties of human language. In formal theories, formalism has a status beyond description, however. It is also the basis of the model of speaker knowledge and must therefore work like a perfect ‘machine’ (efficient, economical and automatic) to generate the correct forms and interpretations. The formal approach therefore necessarily involves a level of abstraction. Although the adoption of an abstract metalanguage and a computational or algorithmic system of rules has certain advantages (it is precise, unambiguous and universally applicable), cognitive linguists (among others) argue that the level of abstraction adopted within the transformational model in fact obscures or misrepresents the reality of human language. Cognitive models of grammar therefore avoid the use of abstract symbols and rules on the whole, although we have seen that cognitive models do rely upon a fair number of complex diagrams aimed at representing the links between grammar and cognition.

اخر الاخبار

اخبار العتبة العباسية المقدسة

للحفاظ على سلامتها.. متحف الكفيل يشرع بخطته السنوية لصيانة مقتنياته

شعبة التوجيه الديني تنظّم مسابقةً ميدانية بذكرى هدم قبور أئمة البقيع (عليهم السلام)

العتبة العباسية المقدسة تواصل إحياء ذكرى هدم قبور أئمة البقيع (عليهم السلام)

قسم الشؤون الفكرية يعقد اجتماعه الفصلي الأول لعام 2026م

المزيد

الآخبار الصحية

اكتشاف هام قد يمهّد لعلاج جديد للسمنة

تغييرات بسيطة في نمط الحياة تقلل خطر أمراض القلب بنسبة كبيرة

ليس اللب فقط!.. قشرة وبذور المانغو تخفي فوائد صحية مذهلة

زيادة استهلاك الكالسيوم ومنتجات الألبان يحمينا من مشكلات صحية خطيرة

المزيد

الآخبار التكنلوجية

العلماء الروس يكتشفون طريقة تحسن فرص نجاح زراعة الأسنان

لحظة زمنية خاطفة في الدماغ تتحكم في التعلم والحركة

القمر لم يعد مجرد محطة.. خطوة جديدة تغيّر مستقبل البشرية

العلماء يكتشفون "نوعا جديد تماما" من الكواكب

المزيد

مواضيع ذات صلة

Who were the creators of early language?

Functions of early language Discussion

Motivations underlying grammaticalization

Cognition or communication?

Functions of early language

Word order

From non-language to language

Grammatical evolution Layers

Early language

On the rise of recursion Conclusions

Generative approaches to grammar

Loss of recursion