Formulation
المؤلف:
Paul Warren
المصدر:
Introducing Psycholinguistics
الجزء والصفحة:
P23
2025-10-29
42
Formulation
The formulation processes are the linguistic powerhouse of language pro duction. Formulation involves grammatical encoding i.e. where the speaker uses their implicit knowledge of grammar to create sentence structures that will convey a message. Chapter 4 examines how formulation also involves phonological encoding, which allows us to construct the appropriate sequences of sound to express the message. Other encoding processes will be involved in producing handwritten or typed text outputs. Grammatical and phonological encoding are linked to different aspects of lexical entries, as shown in Figure 2.6. Words are accessed from the mental lexicon on the basis of their meanings. These content-based or semantic aspects of words are known as lemmas. The lemmas are linked to the forms of words, known as lexemes, which can be the spoken shape of words, or their written forms.
Grammatical encoding Braces conventionally indicate that we are referring to a concept rather than to an actual word. Angle brackets are used to indicate spelled forms of words. Slashes / / are used to indicate aspects of the spoken forms of words. In this text, linguistic examples are also distinguished from normal text by italics or by being indented and numbered. Grammatical encoding is an important aspect of taking the message generated by the conceptualiser and turning it into language. It has been claimed that grammatical encoding involves two separate but related components – functional processing and positional processing Garrett, 1980a . The aim of functional processing is to give the appropriate jobs to words that will express the speaker’s intended meaning. This involves two processes lexical selection and function assignment, i.e. choosing the words and giving them their jobs in the sentence. Initially, abstract forms of words are chosen, based on the concepts that the speaker wants to express. These lemmas are fleshed out into actual word forms, or lexemes, at a later stage. For our example in Figure 2.3, the speaker might select the lemmas hunt, cat and mouse. As we have seen, in terms of thematic structure the cat is the AGENT and the mouse is the THEME of the action conveyed by hunt. Depending on exactly how the speaker wants to convey the message, different possible grammatical roles could be given to these lemmas in the sentence. So for one possible sentence, i.e. the cat hunted the mouse, the AGENT cat is assigned the job of grammatical subject, and the THEME mouse is set up to be the grammatical object. In an alternative sentence expressing the same concept, but with a different focus, i.e. the passive sentence the mouse was hunted by the cat, (cat) is the grammatical object, even though it is still the AGENT or do-er of the action, and the THEME mouse is the subject.
In either of these versions of our example sentence, both the S BJECT and OBJECT concepts are marked as singular and definite which will ultimately lead to the selection of e rather than a as the determiner, so the cat, not a cat . The time of the event being described will be past’. If we had access to the output of the functional component, then we might find there an unordered set of lemmas, together with the functions that they have in the sentence, as in 2.1.
In the next component of grammatical encoding, positional processing, the selected set of lemmas is organised into an ordered string. As part of this process, constituent assembly creates a sentence frame for the message. The details of this sentence frame depend on a number of things, including grammatical considerations such as making sure that tenses are marked using verb inflections, but also specific constraints associated with particular words, as well as the speaker’s decisions about which items to place in focus. A possible sentence frame for our example is given in 2.2, where N stands for noun and V for verb, and determiner’ is a word like a’ or the’.
Next, the lemmas for the content words are accessed from the mental lexicon and slotted into this frame, as in 2.3.
Function words and grammatical endings are then specified, before the process of phonological encoding accesses the word-forms and generates a phonetic plan, which will drive the articulators the speech organs.
Evidence from speech errors
Much of the evidence for the distinction between functional and positional processing as components of grammatical encoding comes from speech errors. Chapters 3 and 4 give more detailed discussions of speech errors and show that errors can provide rich information about language production. Studies of speech errors show that word exchanges like example 2.4 are highly likely to involve two words that have the same grammatical category but which appear in different syntactic phrases. In this case these are the noun phrases that form the subject and object in the sentence. See sidebar for a description of how error data are presented in this book.
We have just seen that word lemmas content-based aspects of words are selected at the functional level of the grammatical encoding system, and are assigned functional roles within the sentence. These roles include grammatical functions that are frequently linked to grammatical characteristics of specific word classes. It is therefore not surprising if exchanges involve two words from the same grammatical category and which could assume the same roles. Let us assume that the abstract pre-verbal message structure of the intended sentence in 2.4 involves concepts in the roles of subject and object of the verb, and let us assume also that at the functional level the two noun lemmas sea t and spring have been selected to express the meanings required of these roles. The error in 2.4 arises when these noun lemmas are assigned to the wrong functional roles, so that when the sentence is subsequently put together at the positional level, they appear in the wrong positions.
Other errors, such as those in 2.5 and 2.6, support the idea that the error in 2.4 results from the wrong function assignment rather than just a superficial misordering of words.
In 2.5 we see that the misplaced items in this case both pronouns have the grammatical marking that is appropriate to the functions they now have, not the functions they should have had. A simple misordering would
have produced the error you must be tight for they. Example (2.6) shows stranding – the plural marker s at the end of the phrase has not moved with the rest of the word Chapter 4 discusses stranding in more detail and shows how this demonstrates that the base forms of words are insert ed into a grammatical framework. Let us assume for this case that the pre-verbal message for the phrase is something like an abstract version of X(singular) full of Y(plural)’. If the functional process has to assign roles and to the two nouns floor and hole and incorrectly assigns these roles i.e. swaps X and Y but not their number marking, then we can understand how we get the error with stranding as in 2.6, rather than a holes full of floor or holes full of a floor.
In contrast, sound exchanges like example 2.7 do not respect the grammatical class of the words they come from, and are most likely to affect words within the same syntactic phrase. Such errors occur at a more local level, during the phrase-by-phrase assembly of the utterance. We will return to a more detailed consideration of sound errors and the construction of the spoken forms of sentences in Chapter 4.
Further evidence that sentence frameworks are developed at least in part independently of the words that are placed into them comes from studies of syntactic priming. One form of this task requires participants to read aloud a sentence the prime before then describing a picture Bock, 1986. Typically, the kind of sentence frame that participants choose for describing the picture is biased towards the structure of the prime sentence. So if participants read the prime sentence in 2.8 and then have to describe a picture of a man reading a story to a boy, then they are more likely to use a sentence like 2.10 than one like 2.11. Sentences 2.8 and 2.10 both use a prepositional phrase to show the recipient, i.e. the phrase starting with to. But if the prime is the sentence in 2.9, known as a double-object construction, then participants are more likely to describe the same picture with a sentence like that in 2.11, also a double object construction.
Importantly, this syntactic priming appears to involve the syntactic structures, rather than the actual words in the sentence. So, for instance the prime sentence in 2.12, which uses a different preposition (for), is just as effective as that in 2.13 in priming the use of the sentence in 2.10 rather than that in 2.11 to describe the picture.
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
