Unit 1 Notes: Descriptive Linguistics, Competence vs Performance, and Data Methods

Descriptive vs Prescriptive Perspectives on Language

• Linguistics is the scientific study of language.
• Descriptive perspective (preferred in the course):
  • Objective recording of how language is actually used.
  • Do not label language behavior as degraded, inferior, or inappropriate.
  • Focus on describing language as used by all people, especially native speakers.
• Prescriptive perspective:
  • Tells people how to speak or write, for various reasons (often related to power).
  • May aim for standardization or improving clarity, but is normative rather than descriptive.
• Course stance: adopt a descriptive perspective for the rest of the semester, describing the language of native speakers.
• Native speakers are individuals exposed to and using a language from early childhood through adulthood; they are central to the description of language.
• We discuss native speakers’ opinions about grammaticality/ungrammaticality of sentences; grammaticality is about what is natural and well-formed within the language, not moral value.

Native Speakers, Language Description, and Grammaticality

• Language to describe: the speech of native speakers; native speakers are those exposed to and using a language from early childhood onward.
• Grammatical vs ungrammatical
  • Grammatical: natural, well-formed expressions according to the language’s rules.
  • Ungrammatical: not part of the language as used by native speakers.
• Colloquial language versus formal, careful speech
  • Colloquial or profane language can be perfectly systematic and rule-governed.
  • Formal vs informal styles may reflect different registers, not different grammars per se.
  • Studying colloquial forms reveals how language rules apply in real usage (e.g., agreement patterns) and can show how language evolves.
• We stress descriptivism but also acknowledge the value of critical evaluation of language use.

Knowledge, Mind, and the Goals of Linguistics

• The cognitive-science angle: understand the knowledge that underlies linguistic abilities.
• We often cannot directly access knowledge inside the mind (brain, thoughts).
• How do we study language knowledge?
  • By observing behavior (speech, sign, writing) as external evidence of internal knowledge.
  • Behavior is influenced by many factors beyond knowledge of language.
• Distinction: competence vs performance
  • Competence: knowledge of the language in the mind (linguistic knowledge).
  • Performance: actual language use (speech, writing, etc.) that reflects knowledge but is shaped by many factors.
• Important caveat: performance data can be distorted by memory, hesitations, false starts, fatigue, intoxication, and other factors.
• We must filter performance data to avoid drawing incorrect generalizations about the underlying knowledge.

Competence, Performance, and Tacit vs Conscious Knowledge

• Linguistic competence: knowledge that speakers have in their mind; set of rules enabling them to generate grammatical sentences.
• Mental grammar: the rule system that underlies competence; used to produce language.
• Linguistic performance is the observable expression of that knowledge, but it is not a direct readout of competence due to performance factors.
• Conscious vs tacit knowledge:
  • Conscious knowledge: can be verbalized, explained step-by-step, and taught (explicit rules, metalinguistic knowledge).
  • Tacit knowledge: knowledge that is implicit, holistic, and not easily verbalized; difficult to introspect.
• Examples and contrasts:
  • Conscious knowledge: you can explain some grammatical rules to someone else.
  • Tacit knowledge: you may have a sharp sense that a sentence sounds wrong but cannot articulate the exact rule responsible.
• Tacit knowledge tends to be robust and often more resistant to forgetting than explicit knowledge (e.g., some language knowledge persists even when explicit rules are forgotten).
• Analogies and related domains:
  • Facial recognition is a complex, tacit, holistic capability similar in some ways to language knowledge; not easily verbalized.
  • These examples illustrate that certain cognitive abilities are built into the brain and operate largely without conscious access.
• Implication for linguistics: aim to study linguistic competence (the knowledge), not just what people can consciously state about language.

Knowledge in the Mind: Conscious vs Tacit, and Implications for Study

• Conscious knowledge (explicit):
  • Can be verbalized and taught; you can slow it down, break it into steps, and explain it.
  • Examples include formal grammar rules you might memorize.
• Tacit knowledge (implicit):
  • Holistic judgments about language; not easily broken down into explicit rules.
  • Robuts and robust across time; often not taught directly.
• The goal in linguistics is to infer the underlying knowledge (competence) from observable performance while accounting for performance distortions.

Data and Methods for Studying Language

• Corpus data (corpora):
  • Large collections of texts or recordings used to observe language use.
  • Can contain billions of words, providing broad exposure to language behavior.
  • Limitations: corpora may include performance errors, typos, or non-representative samples; missing data can bias conclusions.
• Grammaticality judgments (judgments):
  • Native speakers’ intuitions about whether a sentence sounds grammatical or natural.
  • Useful to calibrate corpora findings and correct surprising results.
  • People may later realize a previously judged sentence was a mistake or a slip, providing additional evidence about knowledge.
• Experimental grammars and judgments:
  • Present linguistic expressions to speakers and record their reactions (grammaticality, ambiguity, meaning).
  • Example: judgments about invented word sequences to test sound-structure constraints in a language (
    e.g., which sequences could be English words).
• Ambiguity judgments:
  • Some sentences have more than one possible meaning (meaning ambiguity).
  • Example: "they fed her baby carrots" can mean:
  • Reading A: They fed carrots to her baby.
  • Reading B: They fed her baby carrots.
  • Native speakers know these ambiguities and can report them; prosody and punctuation influence interpretation.
• Data sources complement each other: corpora provide large-scale usage patterns; judgments provide sense of acceptance and meaning; experiments test specific hypotheses.

Case Study: Colloquial Contraction and Dialect Variation (Wanna)

• Phenomenon: contraction of want and to into wanna (e.g., "I wanna go to the movies").
• Rules and variability:
  • In many dialects, want + to can contract to wanna when they appear adjacent:
  • Example: "I want to go to the movies" → "I wanna go to the movies" (informal).
  • Not universal; some speakers/dialects accept contractions in some environments but not others.
• Constraints on contraction:
  • If the clause is in a form like "he wants to go to the movies" → cannot contract to "he wanna go to the movies" for many speakers because there is a phonological/grammatical boundary that prefers the contracted form in certain contexts and not others.
  • In questions like "Who do you want to visit at college?" vs "Who do you want to visit your class?", contraction can shift meaning:
  • "Who do you wanna visit at college?" is acceptable for the first question, asking about the visitor.
  • For "Who do you wanna visit your class?" some speakers find it odd or unacceptable, as it may sound like you are asking about the visitee instead of the visitor.
• Implications:
  • Contraction is not merely sloppiness; it reflects formal conditions on the rule, which differ across dialects.
  • These observations support the view that language contains systematic, rule-governed variants that vary by context and speaker group.
• Ambiguity with contraction:
  • Removing the contraction can change the interpretation of a sentence and alter which reading is intended.
  • Example: a bare version of the question ("Who do you want to visit?") is ambiguous between two readings; adding contracted form can influence or constrain readings.
• Important takeaway:
  • Descriptive linguistics must account for such variation without assuming one standard form is correct for all speakers; rather, rules may be dialect-specific and context-dependent.

Ambiguity and Meaning in Sentences

• Some sentences are ambiguous: they can have more than one meaning depending on structure, stress, or syntax.
• Example: ambiguous pronoun placement and prepositional phrases can yield different interpretations.
• Native speakers’ judgments help identify the intended or possible meanings and highlight where ambiguity arises.
• Prosody, intonation, and punctuation often resolve or shape interpretation, but written forms may leave some ambiguity.

Practical Implications for Descriptive Linguistics

• The descriptive approach guides the development of grammar theories based on actual language use rather than idealized norms.
• We use a combination of data sources (corpora and judgments) to infer the structure of mental grammars.
• We acknowledge performance factors and strive to filter out data that do not reflect competence.
• The study of language intersects with cognitive science: understanding how the mind represents language, how knowledge is stored, and how it manifests in behavior.

Key Takeaways

• Descriptive linguistics records and analyzes language as used by native speakers, avoiding moral judgments about language forms.
• Competence (linguistic knowledge) versus performance (observable language use) are distinct but related; performance is influenced by multiple factors beyond knowledge.
• Much of linguistic knowledge is tacit (not easily conscious); conscious knowledge exists but is not the whole story.
• Corpora provide broad data about language use, but require careful interpretation and correction for performance errors.
• Grammaticality judgments from native speakers are essential to validate theories and to identify subtle grammatical patterns.
• Colloquial language and contractions (like wanna) illustrate that language contains routine, rule-governed variants that differ by dialect and context.
• Ambiguity is a central phenomenon in language, revealing how structure and meaning interact; judgments about ambiguity aid theory-building.

Mathematical and Conceptual Formulations (in-text summaries)

• Competence and Performance relationships:
  • Let K denote linguistic competence (knowledge in the mind).
  • Let P denote linguistic performance (observable language use).
  • A simple model can be written as:
  • $P = f(K, E)$
  • where E represents performance-influencing factors such as memory, fatigue, distraction, intoxication, etc.
• A linear approximation for illustration (not a claim about exact form):
  • $P \,\approx\, aK + bM + cF + dD$
  • where M = memory-related effects, F = fatigue, D = distraction, and a, b, c, d are weights capturing contributions.
• This formalization captures the core idea that performance is shaped by the underlying competence plus situational factors, guiding data interpretation.