Linguistics Lecture Notes: Syntax, Morphology, and Parts of Speech

Syntax, morphology, and the study of parts of speech

• Introduction: Today’s focus is a fundamental, slightly unorthodox approach to parts of speech using a syntactic and morphological lens. The speaker emphasizes that many learners haven’t previously engaged with parts of speech as linguists do, and that the unit will consistently apply this method.
• Key aim: classify words into syntactic categories (parts of speech) to understand how language is structured, not just to label words with traditional school definitions.
• Core idea: syntax is the study of the principles governing how words form sentences and phrases; it’s a descriptive rather than normative grammar.
• Distinction between syntax and morphology:
  • Syntax (grammar in a descriptive sense): how words combine into phrases and sentences.
  • Morphology: study of word formation and internal word structure (prefixes, suffixes, affixes).
• Famous intuition about syntax: structure beyond meaning. Noam Chomsky’s colorless green ideas sleep furiously demonstrates that a sentence can be syntactically well-formed but semantically odd or meaningless.
  • The original order: colorless green ideas sleep furiously (grammatical, meaningless).
  • Reversing the order (furiously sleep ideas green colorless) is not grammatical; it lacks natural English phrase structure and intonation.
  • Intonation: natural English uses phrase and sentence-level intonation that signals syntactic grouping; a shopping-list-like intonation appears when words are not properly grouped into phrases.
• The role of intuition in linguistics: intuitions about acceptability are used as data, not as fixed rules; these intuitions can be analyzed to reveal underlying principles.
• Why syntax matters: it reveals structure in language and informs our understanding of human language knowledge and psychology.
• Morphology vs syntax in practice:
  • Morphology looks at the internal form of words and what kinds of affixes they take.
  • Syntax looks at how words participate in sentences and how they distribute across positions and functions.
• Core concept: a robust theory should generate all and only the well-formed forms (good forms) and should not generate ungrammatical forms (bad forms).
• The lecture’s plan: use morphological and distributional criteria to identify parts of speech, rather than relying solely on intuitive meanings.
• The risk of relying on meaning alone: semantics can be misleading for categorizing words across languages; examples from Irish Gaelic show different syntactic resources can express similar meanings.
• Irish Gaelic example:
  • English: “John is not a doctor.” with doctor as a noun.
  • Irish Gaelic: “John doctors not” (literally), using a verb to express not being a doctor.
  • This illustrates that lexical meaning alone does not determine syntactic category; languages can express similar meanings with different parts of speech.
• Notional vs form-based definitions of nouns:
  • Notional (notional/folk) definitions: e.g., person, place, thing; actions as verbs; adjectives as describing words.
  • Problems with notional definitions:
  • They are broad and insufficient to determine category across languages.
  • Some words that are nouns don’t fit simple “notion” lists (e.g., words without obvious referents or with irregular behavior).
  • They don’t explain why some items on the list are nouns while others with similar meanings are not.
  • The gap: meaning alone cannot determine parts of speech; form and distribution provide more reliable cues.
• The move to a structural approach:
  • Define syntactic categories by morphological form and distributional behavior (where they occur and what they can combine with).
  • This approach aims to capture the brain’s organization of language and its universal properties across languages, not just English.
• The implications of a structural approach:
  • Aims to explain why some forms are allowed and others are not, regardless of meaning.
  • Distinguishes between what is permissible in word formation and what is permissible in sentence structure.
  • Supports cross-linguistic generalizations (e.g., how categories are realized differently in different languages).
• Morphology first: identifying noun-like and other word classes by their word-formation patterns.
  • Example collections: kindness, darkness, wildness, madness; transformation, accusation, rationalization; resentment, happiness, sincerity, certainty, devotion, patient, expectation, specialist, attendant, shrubbery.
  • Observation: these endings −suffixes appear at the end of words, signaling a morphological affix typical of nouns (suffix-based cues).
  • The hyphen in the slide indicates affixes (not prefixes) within a word; i.e., a hyphen marks the boundary of a suffix acting as a morpheme.
• Noun suffixes as strong evidence:
  • Noun-forming suffixes provide a reliable cue that a word is a noun.
  • However, many nouns lack obvious nominal suffixes, so other criteria are necessary.
• An important nuance: some nouns do not show pluralization or other typical noun morphology (mass nouns).
  • Examples: furniture, baggage (mass nouns) do not take plural -s; you cannot say furnitures or baggages.
  • These are still nouns; lack of plural is not evidence against nounhood.
• Tests beyond morphology for nounhood:
  • Possessive/genitive test: mass nouns can take possessive form (e.g., furniture’s price; furniture’s paint).
  • If a word can take an apostrophe-s possessive in a grammatical way, it supports nounhood.
  • Not all nouns pass the same tests; some nouns lack productive morphology or pluralization, so distributional criteria become essential.
• Distributional criteria as a robust method:
  • Focus on how words distribute in phrases and sentences, i.e., their syntactic environments.
  • Determinants test (a strong, classic noun test): determinants tend to occur before nouns (e.g., the car, a bus, this table, these tables, that assassination, those assassinations).
  • A word x that can combine with a determiner and produce a grammatical noun phrase is strong evidence that x is a noun.
  • Important caveats:
  • The test is not bidirectional: not all nouns pass the determiner test (e.g., proper names like Mary typically do not take determiners, as in “the Mary”).
  • The determinant test is one arrow in a multiple-test approach; it does not imply that if something is a noun, it must take determiners.
• The one-way arrow principle in testing:
  • The rule is not bidirectional: an item passing the determiner test is evidence for nounhood, but not passing it does not prove non-nounhood.
  • Similarly, a noun may not pass the determinant test if it’s a proper noun or certain nominal phrases.
  • Other tests must be used in such cases.
• The plural test and other distributional tests:
  • The plural test (ability to take a plural) is strong evidence for nounhood but not universal (mass nouns may lack plural forms).
  • If a word can be pluralized, that supports nounhood; if it cannot, it does not refute nounhood (mass nouns example).
• Adjectives and verbs: distributional tests for other parts of speech
  • Verbs: can take tense. Example: seem → seems, seemed; this tense inflection indicates a verb.
  • Adjectives: can occur between a determiner and a noun (e.g., the clever student; the short book). A word that fits in this slot likely behaves as an adjective.
  • Adverbs: often modify verbs and adjectives; many adverbs end in -ly (e.g., slowly, quickly). Adverbs can often be tested by their ability to modify a verb (John eats slowly) or by their behavior that interacts with the verb.
  • Prepositions: a closed class; basic behavior is syntactic, often followed by noun phrases (e.g., of Mary). Prepositions themselves tend not to have productive morphology and are identified by their distribution.
• Examples and tests in action:
  • Determiner test examples: the clues, a clues, those clues; determiner + noun pattern presence reinforces nounhood; Mary does not take a determiner, so it is not simply classified as a noun by this test.
  • The short book vs the slow book: adjective position test helps identify adjectives, with adjectives typically occupying the slot between a determiner and a noun; “short” vs “clever” demonstrates how position serves as evidence for categorization.
  • Inflection as a verb test: if a word can take tense, it’s likely a verb; e.g., seem → seems/ seemed.
  • Modals: testing with can/should helps identify verbs (e.g., talk → John can talk; talk in “the talk was very boring” is a noun, illustrating the test’s limits and the need for multiple criteria).
• Important practical takeaway about testing:
  • Use a range of tests (morphology, distribution, semantics) because no single test is universally reliable.
  • If a test fails for a word, it doesn’t necessarily mean the word isn’t a noun; the test might not be appropriate for that word.
• Prepositions revisited:
  • There are few productive morphological tests for prepositions; distributional tests (e.g., what follows prepositions: noun phrases) are more informative.
  • The example with “of” followed by a noun phrase (Mary) demonstrates typical prepositional behavior.
• Summary of the approach:
  • Start with morphology to find suffixes and other affixes that signal nounhood.
  • Use the possessive/apostrophe-s test to further validate nouns.
  • Apply distributional tests (determinants, pluralization, position in NP slots) to confirm nounhood, and to distinguish adjectives, verbs, adverbs, and prepositions.
  • Recognize that language-specific variation exists; cross-linguistic evidence (e.g., Irish Gaelic) shows that similar meanings can be realized with different syntactic categories, underscoring the limits of meaning-based classification alone.
• Practical implications for exam preparation:
  • Be fluent with terminology: syntactic categories (nouns, verbs, adjectives, adverbs, prepositions, determiners, conjunctions).
  • Be able to explain why a given word is categorized one way or another using a combination of morphological, distributional, and, when needed, notional criteria.
  • Be prepared to discuss why some words defy easy classification and how different tests contribute to a robust analysis.
• Ethical/philosophical angle: linguistics treats language as a structured system, not a collection of rote lists. The goal is explanatory power and cross-linguistic generalization rather than prescriptive norms.
• Connection to broader language science:
  • The approach aligns with general linguistic theory that structure can be analyzed independently of meaning.
  • Morphology and syntax work together to explain how language is built and how knowledge of language is organized in the mind.
• Final takeaway: a sound, test-driven approach to parts of speech relies on morphology, distribution, and cross-language evidence to identify syntactic categories, recognizing the limits of any single criterion and the value of multiple, corroborating tests.
• Note on notation used in the lecture:
  • The asterisk in a sentence indicates ungrammaticality (e.g., *open un, *the Mary).
  • The transcript illustrates the idea that some forms are allowed and some are not, guiding the development of principles for syntax and morphology.
• Quick reference examples from the lecture:
  • Colorless green ideas sleep furiously: grammatical but meaningless; used to illustrate structure beyond semantics.
  • Unopened: analyzed via suffixes and morphological structure; asked why forms like head open un aren’t possible.
  • Notional vs structural criteria: notional definitions (person, place, thing) are intuitive but not sufficient; structural criteria offer a more reliable basis for categorization.
  • Irish Gaelic example: John doesn’t eat vs John is not a doctor; same meaning expressed with different syntactic categories, underscoring cross-linguistic variation and the insufficiency of meaning alone for classification.
• Overall takeaway for exams: develop a flexible, multi-criterion toolkit (morphology, distribution, and cross-linguistic evidence) to identify syntactic categories, while acknowledging exceptions and the limitations of any single criterion.