PLEASE KEEP REFRESHING YOUR PAGE
The definitions in this current, ‘official’ MCF encyclopedic glossary are those currently implemented by Sharwood Smith and Truscott: a number have changed and a few added since the original glossary was posted online. Since there is some room for different kinds of implementation, other users of the MCF involved with particular aspects of the architecture may want to adjust or add some definitions to fit in with their particular theoretical perspective. If so we would appreciate hearing about them.
Note that there is, especially for linguists, also a downloadable glossary, at the Cambridge University Press website (here), which is specifically designed for readers of the book (The Multilingual Mind: a Modular Processing Perspective) when the framework was called MOGUL. In 2013 the two glossaries were virtually identical. However, as research carried out within the framework proceeds, the current glossary has been and will be continually updated to keep pace with it and so, in various ways, the contents of this version will inevitably differ from what is in the original version: some definitions may change, new entries will appear and old entries may be rewritten.
Users of the MCF framework are welcome to contribute their own suggestions for inclusion in the glossary via the Comments section on the website or directly to this email address.
A B C D E F G H I J K L M N O P R S T U V S W
acquisition: see growth
Acquisition by Processing Theory (APT) (►attrition, resting level, processing)
APT is a specific claim made within the MCF and is supposed to be one principle that applies to all instances of cognitive development. Growth is the lingering effect of processing. The specific nature of the growth is determined by the principles of the modular system in question. With regard to the effects of frequency on the growth of a representations in a given modular system, only processing within the system in question, will trigger growth there. This means that input frequency has to do with the input to that particular system within the mind as a whole and does not refer to current external input from the environment which may or may not or may not get to influence the internal system in question. In other words it is a matter of internal input frequency and not the frequency of activity in the physical neighbourhood of the individual.
activation (► store, resting level/RLA, schema, working memory)
Structures (aka ‘items’ in a store or ‘representations’) are activated to various degrees during on-line processing. They regularly compete with one another as their processor tries to ‘make sense of’, in other words respond coherently to – input and arrange and rearrange them in ways compatible with the principles of the system in question. The state of activation may be taken as a definition of working memory. First and foremost, ‘working memory’ therefore describes activity in individual stores (visual working memory, syntactic working memory, motor working memory, conceptual working memory, etc.) Typically many representations are activated together in a store and in this active state are more or less activated. In other words activation is a matter of degree. This can vary a lot. Activation is the basis for explaining many important facts about the mind (and brain). It is where change (development, learning and loss or attrition) takes place. Furthermore, conscious experience depends upon very high levels of activation ranging across many systems.
affective structure (AfS) (►appraisal, conceptual structure, hub, value)
Affective structures (AfS) in the affective system are what underlie emotions, drives, urges as well as basic positive and negative evaluations that, to a greater of lesser extent, we share with other species. These are processed in the affective module and are conventionally represented in the MCF be being enclosed by exclamations marks: !disgust!, !fear! !pos(itive)! and so on and the most important elements are the primitives which allow negative and positive value to be associated with other representations outside the affective system, influencing current activation levels at any given moment and hence both current processing outcomes and longer term development as well. The affective system is understood to deal with basic emotions and urges that are also shared by other many other species. Complex emotions like pride and embarrassment (shared by perhaps no, or else some other species) are therefore understood to involve external associations, namely with structures in the the (human) conceptual system. The assumption is also that the emotions, basic or complex, will all contain one or other of these value representations. The fact that value associations are formed with many different types of representation means the the affective system, and in particular the affective store acts as one of the two major hubs in cognitive processing.
affective self (►affective structure, conceptual structure, goal-based self, self, schema, value)
A large, semi-structured, affective (AfS) representational schema consisting of various component (affective) representations that correspond to basic drives, urges, motivations and emotions. It is perhaps the most powerful influencer of who we are and how and why we behave the way we do.
A
appraisal (►affective structure, conceptual structure, value)
Appraisals, by which the mind assesses the value of things, are handled at the most basic level in the affective system and at higher levels by the conceptual system interacting with the affective system. The same can be said of emotions which are also handled by these two modules in a similar way. The more primitive emotions in the affective system and the more complex ones partly located in the conceptual system but again interacting with representations in the affective system.
APT (►Acquisition by Processing Theory, processing)
attention (► activation, consciousness)
Attention does not exist as a separate entity but is a convenient term for the outcome of relatively intensive processing of any or all of the perceptual output structures (POpS). It can operate at a subconscious level so not automatically implying consciousness. However, if the activation levels of affected structures rise beyond a certain threshold of intensity they can indeed participate in conscious awareness (see also AfS) but experiments on blindsight, for example, show that we can attend to things without being aware of them.
attrition (► Acquisition by Processing Theory)
This amounts to what is experienced as a relative loss of accessibility of a given representations/representational schemas. The most obvious cause is infrequent participation in the assembling of representations online. Highly inaccessible items appear effectively to be ’lost’ or ‘forgotten’ hence giving credence to a ‘use it or lose it’ principle. Since they regularly fail to get selected, they never appear in overt performance (see also ‘resting level’ although they may prove their continued existence by reappearing later, for example at a later stage in life as a result of the ageing process.. This is Attrition by Processing at work, that is, in some sense ‘APT in reverse’ but there are nevertheless important differences between the two ‘APTSs’.. Competition can be an important factor as well. During processing, new, frequently used representations can come to outcompete old ones such that they become highly active even when the old representation are still more appropriate in the context. Adding a new sport to your repertoire, for example, one that requires similar but not identical movements and strategies to a sport you already practise, can lead to unintended changes in the way you play the first sport as well as transfer effects from the old sport to the new one. Take another example, language attrition: this almost never happens in a vacuum. In particular, loss of resting levels will cause representations associated with the attrited language to be outcompeted by structures associated with other more recently acquired languages. This means patterns observed in attrition are not explainable purely and solely by low frequency of use.
auditory structure (AS/ASL) (► perceptual output structures)
Auditory structures (AS) are the output of the module that processes incoming acoustic input from the environment. They provides input to , for example, the two language systems, specifically PS, but only where the phonological processor is able to build phonological structure as a result. AS, when intensely activated, support the individual’s experience of any type of sound: this includes sounds that are not triggered by input from outside, in other words it can cause experience of sounds that are imagined, sounds in dreams and auditory hallucinations. AS belong to one group of structures provided by the the perceptual system and so are a type of POpS.
awareness (► attention, consciousness, global working memory, perceptual output structures)
B A C D E F G H I J K L M N O P R S T U V S W
blackboard (►activation, memory, store, working memory)
C A B D E F G H I J K L M N O P R S T U V S W
coindexing (► hub, index, interface)
Coindexing is about marking associations, a process by which associations are created between representations in different stores. Representations having the same index are said to be ‘co-indexed’: in MCF terminology, this relationship between the two structures in question is indicated by means of subscript numbers or letters (representation A1 , representation B1). When building a complex representation involving representations in more than one store, the interface between the relevant systems establishes a given index (customarily expresses as some number) on representations to be associated with the result that they activate each other during processing. If A is activated, anything that shares an index with A in other stores, like B, will also be activated. In processing, indexes first have to be created for given representations after which they will continue function as described above and then coindexed is tantamount saying ‘will automatically be coactivated’. Representations also get associated with fellow representations in the same store thus forming more complex representations all written in the same, shared code, the code of the system in question but coindexing is only used for associations between representations across some interface, so ones that do not share the same code.
competition (►activation, resting level)
Along with activation, competition is a major feature of all human processing. When comprehending or producing an utterance, for example, the ultimate ‘solution’, i.e. the meaning we extract from the ‘incoming’ utterance or, alternatively, the utterance we finally produce to express our intended meaning, is only achieved after many competing candidate structures have been reduced to a single sequence of structures. As you peer into the distance at an oncoming pedestrian to see if you recognise them, the same kind of competition between candidates will be taking place as that person gets nearer and nearer. The resolution of any competition is an interaction within and across all the relevant working memories during on-line processing and the current resting levels of all the various items that have been activated. Effectively, if not literally, the system as a whole (the mind) is always looking for a best-fit solution in the circumstances. Given the fact that the world outside is in constant flux, varying all the time, and providing many unclear signals for the mind’s processing machinery to interpret, it makes sense to have a mind (and a brain) that overgenerates candidates in this way simultaneously placing many options at the ready in any situation. What might seem inefficient at first glance, ends up being ultra-efficient.
conceptual structure (CS) (► hub, metalinguistic).
Conceptual structures may be thought of as elements of meaning. This includes areas of meaning that in linguistics are traditionally covered by the terms ‘semantics’ and ‘pragmatics’ and in psychology as a whole by the more comprehensive term ‘semantics’. Meanings, although often expressible by words in a given language, cover all meaning and so do not necessarily have to be linked up with the two language systems in MCF (covering PS and SS). The sound of wind blowing through trees for example, has a meaning while it may or may not have a special word to describe it in your language(s). Conceptual structure is also referred to as CS. Conceptual structure has been also dubbed the language of thought or ‘mentalese’ (Fodor 1975, Pinker 1994): we think (cogitate) in CS: even though the parallel activation of language structures is a typical feature of thought processing as well as many many other types of structure, the actual content of thought is indeed located in the conceptual system. That said, even though we are unaware of the structure of CS, its contents may be made manifest in’ our conscious experience in the form of POpS, ie. as perceptual structures. In this way, can be consciously manipulated. This means that we can go on to analyse the contents of CS in great detail and build up detailed knowledge of its properties. Knowledge about the world we live in is built out of CS. In the MCF, conceptual structures are conventionally represented in capital (block)letters: for example, the concept ‘dog’ is (CS) DOG. The conceptual store ends up acting as a major hub since, when activated, a CS will activates many more associations in other systems, not least the perceptual systems. CS DOG for example will activate visual and auditory associations in the form of VS and AS. In this way, pathways of association in the form of interfaces radiate out from or pass though the conceptual store making it into a major crossroads.
consciousness (►activation, attention, conceptual structure, global working memory, metalinguistic, meta-self, perceptual output structures)
The phenomenon of conscious awareness in the MCF is defined in terms of activation levels: When a representation reaches very intense activation levels, it becomes the object of consciousness. At the lower end of this intensity zone you have attention. Attention does not absolutely require awareness so we are able to notice things with realising it. As the intensity increases so the likelihood of conscious awareness increases too. These extreme levels arise naturally in the stores of the richly interconnected perceptual system, effectively now temporarily merged online: conscious operations are carried out in this Global Working Memory (GWM, the framework’s ‘global workspace’). To become conscious of a meaning, for example, say the meaning of ‘dog’, one cannot be directly aware of the conceptual structure (DOG) that underlies this concept, what it is composed of and how it is being manipulated, only its content. The content comes to us in perceptual form. Conscious experience is also associated with highly synchronised activity across a broad range of systems (and hence, in the brain, a broad physical area as well). The MCF as always reflects current trends in cognitive science, in this case ideas about consciousness.
conventions (► MCF conventions)
core language system/core linguistic system (►language module, Universal Grammar)
This is the mental system that deals exclusively with linguistic structure and is realised in the MCF as a a pair of systems namely the phonological module module and the syntactic module together with their shared interconnecting interface plus two other interfaces that link up these two to three more modules that are not specifically linguistic but do play a major role in language processing, namely the auditory and visual systems (connecting up with the phonological system) and the conceptual system that connects up with the syntactic system. The visual/phonological interface is required for the processing of sign language. Other ways of conceptualising the system or systems dedicated exclusively to language are of course possible. The one illustrated here just happens to be the current preferred version. The terms ‘linguistic processing’ should be generally used in th MCF to refer exclusively to the two systems that handle linguistic structures (PS and CS) while language processing is a much more comprehensive terms involved many different systems that are not themselves ‘linguistic’.
crosslinguistic influence (CLI) (►attrition, transfer)
A generic term for a range of different effects attributable to an interaction between languages systems particularly as reflected in the form of different network of associations within and across the phonological and syntactic systems which can be identified as belonging to different languages. This will include, for example, various forms of what people have often chosen the metaphor of ‘transfer’ between two languages as manifested in language learners’ performance, also changes in an immigrant’s use of the mother tongue due to influence from the dominant language in the community (so-called language attrition ) and even conscious avoidance behaviour by learners trying in their own minds to keep different languages as distinct as possible or a perception of areas of difficulty to avoid in speaking or writing in a non-native language.
D A B C E F G H I J K L M N O P R S T U V S W
development: see growth
E
cement (►primitive, structure)
emotion (►affective structure, appraisal, value)
extramodular (►’language module’, metalinguistic)
This is a term to be used with caution. It is employed only in MOGUL/MCF publications in the context of discussions specifically about language processing. Here, it indicates any aspect of processing located outside what became generally referred to as the ‘language module’ or ‘language faculty’. In an MCF context, this actually means one or other of the two linguistic systems, in effect those that deal with phonological and syntactic structure (although other versions of the language faculty could in principle be integrated into the MCF). Extramodular processing in this restricted sense will involve cognitive structures in stores outside the two linguistic systems – including those immediately related via direct interfaces, for example, stores containing auditory, visual and conceptual structures. For example, the auditory processing of an auditory structure (AS) would be extramodular in this restricted sense if it happened be associated with a phonological representation. By the same token, a conceptual (‘meaning’) representation (CS) that happens to be linked (co-indexed) with a syntactic structure (SS) would also be extramodular as well. Note that this means that , explicit knowledge about the structural properties of a language of which we can be consciously aware, must also be regarded as ‘extramodular’ and strictly speaking non-linguistic in the MCF because it is built out of conceptual structure, and conceptual structure is, despite its rich connections with the two linguistic systems, not in itself linguistic.
F A B C D E G H I J K L M N O P R S T U V S W
faculty (► language module, module)
The term ‘faculty’ refers to a group of modules that together serve a given function, such as language, vision, or music, and therefore tend to show a high degree of coordinated activity, which is to say they are joined together by a large number of well-established schemas. It is a label of convenience rather than an entity of the framework.
feature (► structure, primitive)
Features are representations making up a composite representations within a given theory. The nature of these features is ultimately determined by researchers working in the relevant area of cognitive science, not by the MCF..
G A B C D E F H I J K L M N O P R S T U V S W
GOAL (►goal-based self, goal structure)
GOAL, written with uppercase letters, is a conceptual (CS) representation of the meaning of goal, something that you wish to achieve, an aim, in other words the meaning behind words such as ‘but’ in French, ‘Ziel’ in German, mùbiāo in Chinese and and ‘goal’ in English. This conceptual structure ( CS GOAL) is not itself a ‘goal’ The term ‘goal’ in the MCF, however, is also involved in a particular type of complex representation called a goal structure that is active in goal-driven behaviour. Goal structures in the MCF should be written enclosed with asterisks: *____*.
goal-schema (►goal-based self, goal structure)
goal-based self (►self, affective self, meta-self)
One of three self representations. A semi-structured composite CS representation consisting of a very large number of goal representations. It is the ‘actor’ self.
goal structure (►goal-based self, attrition, schema)
A goal structure, e.g. *seek*, written between asterisks to distinguish for other conceptual representations with the same name, is a primitive CS that is used in a currently active goal schemas.
global working memory (GWM) (►attention, consciousness, working memory, memory, store)
Global working memory (GWM) is associated with conscious processing. There is no global memory store because GWM is a temporary, multi-store phenomenon and it it arises as a result of very high levels of synchronisation and activation affecting representations in the stores belonging to many different systems. The results of this intensification in activation levels are initially subconscious attention then conscious attention and various complex forms of conscious experience.
growth (► APT, attrition, interface)
Also referred to as ‘acquisition’ and ‘development’; the growth of structural combinations within systems and structural associations between systems in response to environmental stimuli that have triggered internal processing operations within those systems. According to Acquisition by Processing Theory (APT), growth is marked by changes in the activation levels of the implicated structures. The shape of growth within a system is controlled by principles specific to that system. Hence, growth in visual structures (VS) is controlled by the principles of the visual system. The same goes for growth in, for example, the phonological system resulting in of new or altered speech representations.
gustatory structure (GS) (►perceptual output structures)
Gustatory structures are perceptual output structures that relate to taste.
H A B C D E F G I J K L M N O P R S T U V S W
hub (►affective structure, conceptual structure)
Looking as the ways in which different cognitive systems interconnect, allowing associations to be made between stores that contain representations that are structured according to quite different principles, two major ‘hubs’ stand out. The first one is the conceptual hub and the second one is the affective hub. In both cases, it would be a mistake to imagine the hub playing some kind of special supervisory role in cognitive processing like a single executive or ‘central processor’. They just happen to function as crossroads where may paths of association intersect. This does however underline the importance of these two system play in cognitive processing and the high level of interconnectivity in the mind in general, also reflected in neural structure.
I A B C D E F G H J K L M N O P R S T U V S W
index (►activation, coindexing, interface, lexicon, resting level)
Indices/indexes are the formal mechanism by which elements in one system’s memory store are chained together (placed in registration with/linked to/associated with) with elements in the memory store of an adjacent module (following Jackendoff). Linguists should not confuse this with the standard use of indices in generative linguistics to mark coreference or lack of coreference. For example, in the MCF, the lexical item ‘meat’ involves the linking of a PS representation /mit/ with an SS representation indicating it is a noun thus: /mit/23 <=> N23 (using the number ’23′ as an arbitrary index). In the MCF, indexes (indices) are treated like representations and can have their own resting level of activation (RLA) influenced by the frequency with which they have participated in online processing.
Each representation has its own activation history determining its resting level. In the current version of the MCF, indexes are understood to have their own resting levels. This is because when particular representations from different stores are coindexed, the resulting association has its own resting activation level reflecting the activation history of that association separate from whatever the individual resting activation levels of the participating representations might be. For example, the colours ‘red’ and ‘green’ will be represented in the conceptual system and both these CS will be associated with the visual representation (VS) of the fruit called ‘apple’ in English . This set of associations will be marked by a common index leading automatically to their regular co-activation. However, if we now assume that a given individual has a stronger association between the visual representation of the fruit in question and the conceptual representation GREEN COLOUR, the index that this CS shares with the visual structure (VS) will have a higher resting level than the resting level possessed by the shared index marking the CS RED COLOUR. As a consequence of this, the ‘green colour’ association will tend to dominate,in other words. be coactivated more strongly than ‘red colour’ even though both red and green concepts will always be coactivated. An increase in the frequency with which red apples are encountered by this indvidual will naturally lead to a raising if the resting level of the index on GREEN COLOUR. and this will change the relative strenth of the colour associations association.
input
‘Input’ is a cybernetic metaphor used to describe what triggers a response from a particular part of the mental system. It is what triggers the response of a particular module to activity in an adjacent system. Each system receives input from one or more adjacent systems. ‘Adjacent’ means a systems is directly connected by an interface with another system.,, Activity within the store of one system will automatically become input to thje store in any an adjacent system . The term ‘input’ in language acquisition studies has long been thought of as environmental input, i.e. the language to which a language user is exposed to and which is supposed to impact on the user’s system in some way. More properly such environmental acoustic activity will constitute input to the relevant perceptual systems. In the case of speech activity, this would be acoustic activity in the environment triggering a response in the auditory system Since the auditory system is interfaced with the phonological system, only the resulting auditory activity will constitute input to the phonological system. In other words this last mentioned system which handle speech structure does not receive input from acoustic activity. Its relationship with speech activity in the environment is indirect because its input actually comes from the auditory system which handles all types of sound input and not just speech and so its input does not directly come from the environment.
interface (►coindex, hub, lexicon, index, schema)
An interface is a simple ‘linking’ processor that exist to connect adjacent systems, or more precisely their stores. Interfaces link structures in different stores so the structures form structural chains of association. These chains often spread to form whole networks, that is to say, representational schemas involving representations in a number of different stores. A given store may have more than one interface but not all stores are directly connected. Indexes are placed on affected representations marking them as associated so that when any one of them is activated, activation will automatically spread to all of the other representations associated, and therefore coindexed with it.
internal context (► affective structure, goal structure, self, value)
Internal context refers to all the internal processing activity taking places and currently influencing the activity within a particular system or systems under consideration. This notion of internal context is in contrast to relevant states and events occurring in the external environment. Internal context consist in factors that fall into two groups:
- the individual’s internally created representations of the external world including aspects related to the immediate situation (current external context).
- inherently internal factors that reflect the individual’s emotions and values as well as their goals and self concepts.
For example, the internal context of linguistic (phonological and syntactic) processing will influence which of the many coactivated linguistic structures actually get used in a particular situation.
J A B C D E F G H I K L M N O P R S T U V S W
K A B C D E F G H I J L M N O P R S T U V S W
L A B C D E F G H I J K M N O P R S T U V S W
language module(s) (►conceptual structure, Universal Grammar)
In MCF in its current form follows Jackendoff in that ‘language module’ – a word to describe the human ‘language faculty- is a cover term for what is actually a bimodular system. That is, it consists of two independent modules, namely the phonological module and the syntactic module. This makes it different from the standard Chomskyan ‘syntactocentric’ conception. Together with their various interfaces, the two systems (handling PS and SS) form what is also called the the uniquely human, language ‘faculty’. Taken as a whole, human language cognition which underlies both internal reflection and external communication with other members of the species covers a much wider area than the language module. However, is this bimodular system which makes human language structure unique. Meanings as expressed via the medium of language, in other words ‘semantic’ and ‘pragmatic’ representations, lie outside the language module. They are conceptual structures (CS) which can cover any meaning whether or not it happens to be readily expressible in one or other of an given individual’s languages have a ready-made. Music and other form of aesthetic experience may for example convey recognisable meanings that are difficult or impossible to express in he form of language.
lexicon (►schema)
There is no general lexicon in the conventional sense, i.e. as a single store. The ‘lexicon’ of a language in this sense is a metalinguistic notion useful for an analytic understanding of language but less helpful in understanding the actual way the mind works. A lexical item can be seen in two alternative but compatible ways: 1) as a rule associating three separate and independent types of structure, a phonological structure (PS) a syntactic structure (SS) and a conceptual structure (CS). In processing terms, it is best thought of as 2) a chain or small ‘schema’ of structures that have been activated and put in correspondence with one another: a phonological structure associated with syntactic structure associated with a conceptual structure, i.e. PS<=>SS<=>CS. Each structure is processed independently within its own processing unit (module) according to the principles of that module. The nearest thing to a lexicon in the MCF is the unique store of structures that each system possesses, hence the phonological processing unit contains a phonological ‘lexicon’ in which only phonological structures are stored, either as independent elements, or typically in various combinations specific to some language or languages known to the individual language user and formed during exposure to that language.
M A B C D E F G H I J K L N O P R S T U V S W
memory (store) (►interface, store, working memory, global working memory)
Memory is normally conceived of as a store of people, things events that have happened and that can be recalled on particular occasions. Each store in MCF also contain some elements that were there at birth, for example combinations of primitive representations sometimes linked across stores to optimise survival chances enabling instinctive responses to certain things and events that might in principle be useful or dangerous. These can hardly be called ‘memories’ except perhaps in evolutionary terms so strictly speaking stores are not ‘memory stores’ as such so much as stores of elements than can be used to form memories. Such memory material in the MCF is modularised. Auditory memory is formed from representations in the auditory store. Conceptual memories are formed in the conceptual store and so on. This means that memory is not to be seen as a ‘common pool’ that different parts of the mental system have access to. This goes both for long term memory and working memory. In fact the distinction between long term and working memory is minimal: it resides in different levels of activation and working memory is nothing more or less than ‘currently activated representations’. ‘Memories’, as used in everyday language, will very often be built out of coalitions of items linked via interfaces across individual stores, a classic example being ‘episodic’ memories which do not exist under a separate rubric in the MCF but which can reflect complex coalitions of memories involving various perceptual structures, conceptual structures and perhaps also motor structures and structures within the language module as well. ‘Global’ working memory is something different and is explained in its own entry.
metalinguistic (►consciousness, POpS)
As with any kind of metacognition, knowledge about language, which can be projected into to awareness via POpS, is created in conceptual structure and not in the two linguistic systems.
meta-self ► affective self, affective structure, conceptual structure, goal-based self)
One of three instantiations of self. A large composite conceptual (CS) representation consisting of the CS ME combined with representations of characteristics attributed to the person: I am X, I am not X. This is what is commonly called the ‘conceptual self’. It probably includes representations of ‘I should (not) be X’, ‘I hope to be X’, ‘I fear becoming X’… (the ought to self, the hoped-for self, the feared self…).
module (► interface, language module, interface, store)
The mind is functionally subdivided into relatively independent modules. In the MCF, these modules are generally referred simply as ( the mind’s) systems. The modular organisation of the mind enables specific functions to be conducted by a given system without any direct influence from outside. Each modular system has its own principles guiding the assembly and arrangement of its representations and consequently structures that are unique to that system. In other words, each module operates with its unique code and unique set of principles. For example, one of the linguistic modules, the phonological module, operates with phonological code and phonological principles that are not shared or translatable into any other code. At the same time the basic structure of a modular system is generic and applies across the cognitive system as a whole. That is, any module consists of 1) a processor and 2) its own store. Stores are interfaced with one or more different modules allowing the formation of co-activated chains or networks of structure. It is assumed that brains also have their modular make-up but a mental or ‘cognitive’ module should be kept distinct from any particular neural subsystem or location: typically, when cognitive and mental systems are aligned with one another, the functions of a given cognitive module will be associated with different regions and subsystems in the brain that cannot be identified solely with one particular region..
Modular Cognition Framework (MCF) (►MOGUL)
MCF is the name for the framework of which MOGUL is the instantiation most discussed far. The difference is only in that ‘MOGUL indicates the focus on the role of language in the mind. The framework however is not a language research framework per se since it is about cognition in general, hence its more neutral title: the ‘Modular Cognition Framework’.
MCF conventions
The way various categories are represented in the MCF may be illustrated as follows, often using the dummy word ‘cat’ as a substitute term preceded by an equals sign and then the abbreviated term.
We remain open to further suggestions from researchers in particular areas of expertise especially regarding the perceptual, spatial and motor conventions that might be applied to different areas of cognition).
Affective structure = AfS !fear! Unchanged: Affective structure = AfS !fear!
Auditory structure = AS: 1) a representation of the sound of the word, cat [khæt])
2) representation of all other types of sound [creaking sound]. Unchanged:
Conceptual structure = CS BEAUTY, CAT, (the concept of a) PREPOSITION
Goal structure *(NAME OF GOAL)*
Gustatory structure = GS
Motor structure = MoS
Olfactory structures = OfS
Perceptual Output Structures: POpS
Phonological structure = PS /æ/
(following preferred conventions)
Primitive = where required to to avoid confusion, indicate primitive status by adding
asterisks: CS *SEEK* (the goal structure) versus CS SEEK (the meaning of ‘seek’)
Resting level of activation= RLA
Somatosensory structure = SmS
(Spatial structure = SpS (the spatial module not yet been incorporated into the MCF )
Syntactic structure = SS N, Adv, Prep, Det, etc.
(following preferred conventions)
Visual structure = VS: 1) = #“cat”# (orthographic VS)
2) #cat sign# (sign language VS)
3) = VS #cat image#
MOGUL (►MCF)
MOGUL stands for Modular Online Growth and Use of Language and was the original name for the underlying theoretical framework. The name now refers to the main focus of project work carried out since 2000 using this framework, i.e. exploring the way language(s) are acquired stored and processed within the mind as a whole. In other words, ‘MOGUL’ is now used as the cover name for this language-focused project. This means that the MOGUL project should therefore be distinguished from the supporting theoretical framework is called the Modular Cognition Framework (MCF).
motor structure (MoS) (► somatosensory structure)
–
N A B C D E F G H I J K L M O P R S T U V S W
O A B C D E F G H I J K L M N P R S T U V S W
olfactory structure (OfS) (► perceptual output structures)
Olfactory structures belong to a group of structures making up the perceptual system called POpS, in this case, having to do with the sense of smell (►perceptual output structures).
offline (► processing)
online (► processing)
orthographic structure ,► VS, perceptual output structures, visual structure)
Orthographic (spelling) structures do not exist independently of visual structures as a whole, i.e. as a separate type of POpS . They are visual structures (VS) that have links with/are interfaced with PS and CS . These associations makes them part of language in the broadest sense. In other words, they are visual structures that happen to have been created specifically as a result of interaction with the language module and as a result of reading experience. In the same way, auditory structures that happen to have been created specifically as a result of interaction with the language module and as a result of listening experience, might be called ‘phonetic’ structures. As far as the cognitive system is concerned, neither orthographic structures nor phonetic structures have any separate status. If it is convenient to identify structures that are either orthographic (spelling-related) or phonetic, this can be done by using italic script enclosed in parentheses #“cat”#.
P A B C D E F G H I J K L M N O R S T U V S W
parsing (►conceptual structure, consciousness, metalinguistic, module, processing)
Parsing is a term used for the structural analysis of stretches of language. In processing, it is the online structural analysis of utterances, often associated with syntactic analysis, but potentially applicable to any kind of linguistic structure. It also tends to imply receptive (comprehension) processing. However, depending on the preferred theory, it can also be applied to the online creation of structure during language production. This more inclusive interpretation is assumed in the current framework. Furthermore, while the term ‘parsing’ is only used in connection with linguistic structure, in this framework it is actually no different in its essentials from any kind of online cognitive structuring process. It refers, in any mental subsystem, to the operations of that module’s processor. The processor attempts automatically to combine according to its own combinatorial principles currently activated representations in its own memory store, favouring first the most highly activated ones. Applicable metaphorical phrases such as ‘analyse’, ‘attempts to combine’, ‘operates on a best-fit principle’ and’ parsing strategy’ may be used to describe such automatic operations as long as they are treated as just metaphors without implying the oversight of any supervisory system. At the same time, the term ‘parsing’ can also be used in its more traditional sense to refer to metalinguistic processing where there is (slow. sequential resource-hungry) conscious supervision. In this case, the most important module involved is the conceptual system working with those conceptual representations (CS) that are sufficiently well established to have become accessible to conscious awareness and form the basis of reflection, planning and other thinking processes. Any lingering effects of such conscious operations that may be considered as changes to an individual’s current knowledge are stored as conceptual representations, i.e. in conceptual memory.
perceptual output structures (POpS) (►auditory structure, gustatory structure, memory, olfactory structure, somatosensory structure, visual structure)
Perceptual output structures is a generic term covering all the structures, like auditory structures (AS) and visual structures (VS), that constitute the output of the perceptual systems. A crucial feature of POpS is the existence of strong interconnections between the individual stores so that, together, they form a well integrated system. This integration, with the very high activation levels that result from it, plays a crucial role in the MCF accounts of awareness and working memory. Of the various possible types of perceptual structure, only five have figured in the framework. This number, as well as the characteristics of POpS, naturally depends on which theory of perception has been selected for instantiation within the framework.
phonetics (►auditory structure)
What might otherwise be called ‘phonetic’ structures in the MCF do not exist independently from auditory structures (AS) as a whole, i.e. as a separate type of POpS. This means that phonetic strings, traditionally represented using square brackets, are actually auditory strings, i.e. auditory structures (AS) that happen to have been created specifically as a result of interaction with the language module and as a result of listening experience (see also the discussion in the section on orthography). If it is convenient to identify auditory structures that are phonetic, this can be done by using IPA phonetic script: [khæt].
phonological structure (PS) (►auditory structure, phonetics)
Phonological structures in the MCF are represented following standard phonological convention in simple phonetic script e.g. /kæt/. More specifically, phonological structures will be represented following the conventions of the particular phonological theory being applied. The storing and processing of phonological structure is carried out within a closed-off expert system or module: even though we can become aware of features of auditory structure with which PS is linked (like stress, rhythm, sibilants, etc.), we remain completely unaware of the structure of PS itself (cf. Jackendoff 1978: 88; its contents are inaccessible to awareness and therefore cannot be consciously manipulated in any direct manner.
primitive (►growth, Universal Grammar)
Primitives are the essential building units that make up representations that are developed over the course of the lifetime. They should themselves be treated as representations but unlike most representations an individual possess at any given moment, they are supplied in advance, i.e. they are innately given. Every modular system has them. The visual system has them because they are needed to assemble visual representations that create our uniquely human visual perception of the world around us. This means the visual primitives of a dog or cat, for example, may well differ from ours as also the principles that combine them since dogs and cats both see the world very differently. They (the primitives) are typically associated within their store with other primitives in many different combinations (composite representations). The principles of combination are determined entirely by the given processor in question. Where there might be cause for confusion, a specific primitive will be indicated by surrounding asterisks as in the conceptual representation. There are two possible ways of viewing primitives in the MCF. One is as presented above as a ready-made representation as part of our biological inheritance. The alternative version would be that a primitive is provided in advance not yet as representation a but rather as a property of a given system’s processor: it is then used, when appropriate, in the construction of a new representation. In both version primitives are by definition supplied at birth for each and every system. They will be species-specific and so, in this case, uniquely human.
processing
In a modular system processing is in essence a local matter, occurring within individual modules. But processing can also be seen at the global, system level, since activity in one module necessarily interacts with that in others, via interfaces. Local processing consists of a processor combining active representations in its store in order to establish coherence there. Because of both internal dynamics and influences from other modules, this coherence is inevitably brief, as additional representations rise in activation level and therefore compete for dominance. At the global level, processing consists of active representations in the various modules influencing one another’s activation levels, via interfaces. When a visual representation of a horse is active in VS, for example, its activity spreads to the concept HORSE in CS, which might then come to dominate its store. The cognitive system as a whole is designed to seek broad, global coherence, in the form of a dominant schema (a horse schema, for example). This broad coherence is achieved when each store is dominated by a node of the schema, each supporting the others through spreading activation. Global coherence, like local, is typically brief, as constant competition occurs among active schemas.
processor (►structure, primitive, processing, schema, store)
The processor is one of the two components of a module, the other being the store with its representations. A processor’s function is to establish coherence within its store, coherence being a state in which a single representation dominates. It does this by combining relatively active representations to produce a composite representation that is consistent with its in-built principles – syntactic principles for the syntax processor, visual principles for the visual processor, and so on. Importantly, this is a purely local, module-internal function. The syntax processor for example can only deal with syntactic representations.
representation (► structure, primitive, processing, schema)
A representation is a structure, simple or complex, in the store of a module, encoded specifically in the terms of that module. A syntactic representation is ultimately composed of syntactic primitives, an auditory representation of auditory primitives, and so on. Each representation has a currrent activation level, interpreted as its current availability to its processor, and an abstract resting activation level. The term ‘representation’ must be understood flexibly because strictly speaking structures do not ‘represent’ anything (Jackendoff 2002: 20).
resting level of activation (RLA) (►activation, schema)
A resting level of activation (RLA) is an abstract property of an individual representation, that is, anything from a single primitive to a highly composite representation. It can be defined as a particular base level of activation, a representation’s current activation level consisting of its RLA plus the effect of any stimulation it is currently receiving. Thus, the higher a representation’s resting level, the better chance it has in any competition. A representation’s resting level rises slightly each time it is used in processing and gradually declines when it is not used.
RLA (resting level of activation) (►activation)
schema (► index, memory, primitive, representation)
A schema is a network of representations extending across different modules, connected by interfaces. The individual representations are the nodes of the schema. The effect of the connections, especially when they are well established (have high resting levels), is that representations of different types, visual and conceptual for example, act like a single unit in processing. A word, like cheesecake, is an example of a schema, consisting of at least a PS, an SS, and a CS, but potentially extending to various perceptual modules (the look, smell, and taste of the cake), motor structures (the pronunciation of the word and the process of eating cheesecakes), and affective structures (the person’s feelings toward the cheesecake). If schemas are thought of as ‘patterns of association between representations’, it is important to distinguish them for similar patterns of association between representations written in the same code within the same store. By contrast, schemas, for their formation, depend on the presence of interfaces that connect stores and enable associations with (and subsequent coactivation of) representations in different stores. (Note that the current definition of schema differs for the original 2008, i.e., one in the first book devoted to this framework – The Multingual Mind).
self (► affective self, goal-based self, meta-self)
A complex notion involved for example in conscious experience and in goal-based processing. In the MCF it is broken down into three separate concepts: the affective self (an AfS) , the meta-self and the goal-based self (both CS).
somatosensory structure ( SmS)(►perceptual output structures)
Somatosensory structures are the output of a group of sensory systems relating to touch, pain, temperature, body position and sense of the body generally.
spatial structure ( SpS) (► module).
SpS is the module responsible for representing space: the shapes, positions, orientations and movements of objects and their relation to each other and to our bodies. It has so far received little development in the framework.
store (►memory, primitive)
A store is one of the two components of a module, the other being the processor. It stores structures of a particular type, some of which will be primitives and others combinations of structures within that store that have been formed through experience. The totality of an individual’s memory is the complete set of all these various specialised stores and the connections between them. Given this modular view of memory, there is, therefore, no single all-encompassing memory faculty.
structure (►index, memory, primitive, representation)
Structure is equivalent to the more common term ‘representation’. It has the advantage of being in a sense more neutral.
subgoal (►goal)
A subgoal is a goal representation that is seen as contributing to achievement of another goal. When a goal schema is active, a subgoal is any relevant goal representation other than the primary goal. In terms of its structure and activity, a subgoal is simply a goal, no different from any other. Any goal can be a subgoal in a given instance and then cease to be one when CS activity changes.
transfer (►crosslinguistic influence)
This is a term conventionally used in language acquisition studies to describe the modelling of structures in a developing language on the patterns of structures in another language known to the learner. Transfer as a metaphor can be misleading given that what is transferred from A to B should actually disappear from location A. Also, not everyone would agree that language systems literally exist in separate locations. In the MCF all languages are served by the same system, and transfer effects are explained in terms of interactions between representations within a common store during processing.
UG /Universal Grammar(►language module, primitive)
In the MCF, UG provides the principles that make up the processors of the linguistic system, consisting of PS and SS, and the primitives with which those two processors work. The precise nature of the principles and the primitives is a matter of whatever linguistic theory has been implemented, as long as they comply with the basic architecture of the MCF. As with any type of primitive, they can be more or less elaborate according to the preferred theory. Thanks to the limitations UG imposes on how grammars may be constructed, any grammatical system is fully learnable by children, given full exposure to the language. Every cognitive system has it own principles that work in a similar manner so that UG is nothing special but rather represents the norm.
value (►affective structure, appraisal)
In the framework, value refers to two representations in affective structures, !pos! and !neg!, that constitute the value, positive or negative, that is assigned to everything we experience. It has its roots in the basic need of every organism to avoid potentially harmful things, like predators, and seek potentially beneficial things, like food and sex. Representations in general are connected to !pos! and/or !neg!, and these valuations play a crucial role in all aspects of cognition and behaviour, including the acquisition and use of a second language. Value is also the foundation of emotion, which is to say that value representations are components of affective representations in general.
visual structure (VS) (►perceptual output structures)
VS constitutes the output module of the vision faculty. In other words its representations are the final product of each instance of visual processing. VS thus belongs to the group of structures making up POpS. It also supports internally generated visual experience, i.e., sights that are imagined or appear in dreams and hallucinations. VS is interfaced with PS and is intimately associated with orthography and sign language.
word (►lexicon)
working memory (►global working memory, memory, store, consciousness, POpS)
Working memory (WM) refers to the way that information, in the form of representations, becomes available for use by processors. The MCF follows the line of thought that sees it as patterns of highly activated structures along the lines proposed by Cowan (1993, 2001) and others, where level of activation corresponds to availability. The crucial difference is that the MCF account is modular. A processor can only use representations in its own store, encoded in the specialized format of that module. It follows that each module has its own WM. Interactions among these local WMs, involving activation of schemas, constitute a global WM, making coherent information (schemas) available to the cognitive system as a whole.
A B C D E F G H I J K L M N O P R S T U S W