What the Speaker Knows: Language, Compositionality, and the Limits of Explanation
Imagine hearing someone say, “The astronaut left a sandwich beside the telescope.” You have never heard this particular sentence before, yet you can understand what happened and how its elements relate to one another. You do not need to have memorised the sentence in advance to grasp its meaning.
This ability is part of our ordinary experience of language. We understand unfamiliar expressions, recognise relationships between words, and notice when a change in formulation alters what a sentence means. We can also distinguish between sentences that contain similar words but organise them differently. Our command of a language therefore extends beyond the expressions we have encountered and learned individually.
What makes this possible? We clearly draw on what we already know about words and their use, but that knowledge must somehow enable us to interpret combinations we have never encountered. Describing the meanings of individual words and identifying familiar grammatical patterns help us understand particular expressions, although a further question remains: how can a finite knowledge of a language support such a wide range of interpretations?
This question invites us to investigate the regularities that make linguistic understanding possible. To do so, we must consider how expressions are organised, what determines the value of their constituent elements, and what a theoretical account of these features can tell us about the ability to use a language.
2. When Understanding Becomes a Problem
Ferdinand de Saussure draws attention to a difficulty that arises when we attempt to analyse language systematically:
“Doubtless these difficulties do not arise for the language-users themselves. Anything which is significant in any way strikes them as being a concrete unit, and they do not fail to notice it in discourse. However, it is one thing to sense this rapid and subtle interplay of units, but quite another thing to give an account of it by means of a systematic analysis.”
The passage invites us to distinguish between recognising significant elements as we use a language and explaining how those elements interact. In ordinary discourse, speakers identify meaningful units without having to reconstruct the relations that organise them. A systematic analysis, by contrast, must make those relations explicit and account for the role they play in linguistic functioning.
The distinction concerns two activities rather than two kinds of people. Linguists are speakers too, and their practical command of a language enables them to recognise expressions and use them appropriately. Yet being able to navigate a linguistic system does not automatically provide an account of how it works. What is familiar in practice may become difficult to describe when we ask how its elements are organised and how their relations contribute to understanding.
Saussure’s observation thus gives us a way to formulate the problem that guides this inquiry. If speakers can use a language without explicitly analysing its organisation, what can a systematic account establish about the regularities that make such use possible? We can begin by examining how the components of an expression contribute to its interpretation, before turning to the relations through which linguistic units acquire their value.
3. Compositionality: How Expressions Acquire an Interpretation
One approach to linguistic understanding focuses on the relationship between the components of an expression and the interpretation of the whole. The principle of compositionality states that the interpretation of a complex expression depends on the interpretations of its parts and the way those parts are structurally combined. It offers a way to account for how linguistic structure contributes to meaning.
Consider the sentence “The musician who lives next door repaired my bicycle.” Understanding it requires us to interpret the phrase “the musician who lives next door”, including the relative clause that identifies the person being discussed, and to understand how that phrase relates to “repaired my bicycle”. The interpretation of the sentence depends on the contributions of its components and on the grammatical relations between them. If the organisation changes, the interpretation may change as well.
Compositionality helps illuminate two features of linguistic competence. Productivity concerns our ability to understand and produce expressions beyond those we have previously encountered. Systematicity concerns the connections between our ability to interpret one expression and our ability to interpret other expressions that share relevant components and structures. These features help explain why knowledge of a language can support an extensive range of interpretations rather than being restricted to a collection of memorised sentences.
The principle addresses a particular question: how do the components and structure of an expression contribute to its interpretation? Identifying this regularity is an important achievement, because it makes the relationship between linguistic form and interpretation available for systematic investigation. Further questions remain, however, about how speakers acquire the relevant abilities and how comprehension takes place in practice. To investigate another aspect of linguistic organisation, we can turn from the interpretation of complex expressions to the values of the units from which they are formed.
4. What Makes a Linguistic Unit What It Is?
Saussure challenges the conception of language as a nomenclature: a collection of names attached to objects whose identities are already established. This conception treats words as labels for things and leaves unexplained what makes them specifically linguistic units. Saussure instead understands language as a system in which the value of a sign depends on its relations with other signs.
The linguistic sign unites a signifier, understood as a mental sound-image, with a signified, the concept associated with it. These two aspects belong together, but their correlation does not by itself determine the sign’s full linguistic value. A sign also acquires its value through the differences that distinguish it from other signs and through the relations that place it within the language.
The distinction between English sheep and mutton and French mouton illustrates this point. English has different words for the animal and its meat, whereas French mouton can refer to either, depending on the context. Although the words belong to related areas of experience, their values do not coincide. The distinctions made within each language help determine what its terms mean in relation to the alternatives available in that language. A word’s linguistic value therefore cannot be established simply by identifying the object to which it refers.
Saussure also distinguishes between two kinds of relations. Syntagmatic relations connect elements that occur together in a sequence, while associative relations link a term to other elements that could occupy a comparable position or belong to a related field of linguistic possibilities. When we use a word, its role in the expression depends partly on the words with which it combines and partly on alternatives that remain absent from what we say.
This account directs attention to the network of differences and relations through which linguistic units acquire their place within a language. It offers a way to investigate why a unit has a particular value in one linguistic system and why that value cannot simply be read off from the thing the unit names.
5. What Does a Linguistic Theory Explain?
The two approaches we have considered direct our attention to different aspects of linguistic organisation. Compositionality helps us examine how the components and structure of an expression contribute to its interpretation. Saussure’s General Linguistics directs us towards the relations through which individual signs acquire their value within a language. Each makes certain regularities available for systematic investigation, although the regularities at issue are not the same.
Both approaches also draw attention to something that speakers need not formulate explicitly in order to use their language. A speaker can understand a complex expression without being able to state a compositional principle, just as someone can use words appropriately without explaining the workings of the Language Mechanism. The regularities identified by a theory need not be consciously represented as rules by everyone who speaks the language.
This raises a broader question about the relationship between theoretical explanation and practical competence. Consider three related questions. What regularity can we identify in the interpretation of expressions? What relations within a language account for the values of its units? And how do speakers acquire and exercise the ability to understand and produce expressions? Each question concerns linguistic functioning, but each requires a different kind of inquiry.
A compositional analysis can clarify how the structure of an expression contributes to its interpretation. A Saussurean analysis can reconstruct the relations through which a linguistic unit has its value. Neither achievement, taken by itself, explains how speakers acquire the relevant abilities or how those abilities develop and operate in practice. Such questions require further investigation into the acquisition and exercise of linguistic competence.
The distinction matters because a theory may explain an aspect of language without accounting for every condition of its use. Identifying a regularity is more than merely observing that something happens: it allows us to describe a pattern, establish relationships and understand how certain interpretations are made possible. Reconstructing the system in which linguistic units function can deepen that account by showing how their values depend on one another. Yet explaining these features of language does not automatically explain how an individual comes to command the system.
The relationship between practical knowledge and theoretical explanation is therefore more complex than either simply equating them or treating them as unrelated. Our practical competence provides the phenomenon that linguistic inquiry seeks to understand, while theories make selected aspects of its organisation explicit. Their contribution becomes clearer when we ask what each account explains, rather than expecting one kind of analysis to answer every question about linguistic understanding.
6. Understanding a Language, Understanding Its Conditions
Return to the sentence about the astronaut and the sandwich. Understanding it required no prior encounter with that particular combination of words. We interpreted it by drawing on our knowledge of English and on the organisation that makes its expressions intelligible.
Theories of language help us investigate the conditions that support this capacity. They identify regularities in how expressions are interpreted and examine the relations through which linguistic units acquire their values. Such accounts clarify important aspects of linguistic functioning, even when they leave further questions about the acquisition and exercise of competence open.
Understanding a language is therefore a practical ability whose organisation can be investigated from different theoretical perspectives. The more precisely we identify what a theory explains, the better we can assess its contribution and recognise which questions remain to be explored.
The question with which we began has consequently acquired a more precise form. We can ask not only how speakers understand expressions they have never encountered, but also what kind of explanation makes that understanding intelligible. Pursuing the question further means returning to the theories and their original texts, examining how they formulate the problems, and deciding for ourselves how much their accounts allow us to understand.
References
Saussure, F. de. (1959). Course in general linguistics (W. Baskin, Trans.). Philosophical Library. (Original work published 1916).
Heim, I., & Kratzer, A. (1998). Semantics in generative grammar. Blackwell.
Comments
Post a Comment