Linguistic Units & Grammatical/Syntactic Categories – Comprehensive Study Notes
Overview: The Architecture of Language
Language is structured hierarchically. Each higher-level unit is built from the units below it, much like bricks → walls → rooms → an entire house.
- Morpheme → smallest meaningful form
- Word → one or more morphemes
- Phrase → word‐group acting as a single syntactic unit
- Clause → phrase-group with its own Subject + Predicate
- Sentence → one or more clauses that express a complete thought
Understanding the properties of, and relations among, these units underpins morphology, syntax, semantics, psycholinguistics, language acquisition, NLP and AI.
Morphemes
A morpheme is the minimal linguistic sign that carries meaning or grammatical function and cannot be subdivided further without loss of meaning.
Main Division
- Free morphemes – stand alone as words.
- Lexical (content/open-class): nouns, verbs, adjectives, adverbs. New items can be coined (blog, selfie).
- Functional (grammatical/closed-class): determiners, prepositions, conjunctions, pronouns, particles. Rarely admit new items.
- Bound morphemes – must attach to a host.
- Derivational – create new lexemes or change word-class: happy → happi-ness, act → re-act, write → writ-er.
- Inflectional – encode grammatical categories without changing class: walk → walk-ed, cat → cat-s. English has 8 inflectional suffixes ((-s, -ed, -ing, -en, -er, -est, 's, -s) for person/number).
Significance
- Explains spelling & pronunciation alternations (heal → health; electric → electricity).
- Essential for parsing, machine translation, spell-checking.
Words
A word is the smallest free unit that can occupy slots in syntax and convey meaning.
- Etymology ((\textit{etymon}+\textit{logia})) traces historical origin.
Structural Classes
| Class | Morpheme Count | Example |
|---|---|---|
| Simple | 1 | kind |
| Complex | >1 | kind-ly, un-kind-ness |
| Compound | ≥2 free stems | tooth + brush → toothbrush |
Major Word-Formation Processes (English)
- Borrowing (guru, ballet)
- Compounding (credit-card)
- Blending (brunch, smog)
- Clipping (demo, lab)
- Back-formation (edit ← editor)
- Conversion/Zero-derivation (to email, to google)
Phrases
Phrase = group of words lacking its own subject–predicate, behaving as a single unit. Key features: no finite verb (except inside an embedded clause), headedness, internal flexibility.
Head & Modifiers
- Head determines syntactic category.
- Pre-/Post-modifiers expand or restrict meaning.
Major Phrase Types (English)
| Phrase | Head | Core Function | Sub-types / Notes |
|---|---|---|---|
| Noun Phrase (NP) | Noun | Subject, object, complement | 8 sub-types: simple, expanded, complex, with pre-/post-modifier, quantifier, possessive, embedded clause, etc. |
| Verb Phrase (VP) | Verb | Predicate | simple, with auxiliary, with modal, negative VP. Up to 4 auxiliaries possible: might have been being…. |
| Adjective Phrase (AdjP) | Adjective | Modifies NP | 7 sub-types incl. comparative, superlative, Adj + PP, Adj + clause. |
| Adverb Phrase (AdvP) | Adverb | Modifies V, Adj or Adv | 8 semantic classes: place, time, manner, degree, cause, agent, condition, comparison. |
| Prepositional Phrase (PP) | Preposition | Adverbial, post-modifier, complement | 8 semantic roles: place, time, direction, manner, reason, agent, condition, comparison. |
Clauses
A clause is a group of words containing (at least) a subject and a finite verb. Larger than a phrase, smaller than a sentence.
Independence
- Independent/Main clause – expresses a complete thought; can stand alone.
- Dependent/Subordinate clause – requires a main clause; functions as:
- Noun Clause – What she wrote is fascinating.
- Adjective / Relative Clause – The man who is wearing red…
- Adverb Clause – Because it was hot, I drank water.
- Coordinate Clause – linked by FANBOYS (and, but, or…).
Comparison Table
| Property | Independent | Dependent |
|---|---|---|
| Completeness | full proposition | fragment |
| Finite verb | yes | yes |
| Introducers | none/coord conj. | subord conj., relative pronoun |
Syntactic Categories (Parts of Speech)
Lexical (Open-Class)
- Nouns – reference; sub-types: common, proper, abstract, concrete, mass, count, collective.
- Verbs – actions/states; transitive vs intransitive; linking vs auxiliary; argument structure.
- Adjectives – modify nouns; gradable, comparative, superlative.
- Adverbs – modify V/Adj/Adv; classes: manner, time, place, frequency, degree.
- Prepositions – mark spatial/temporal/logical relations; head of PP.
Functional (Closed-Class)
- Determiners – articles, demonstratives, possessives, quantifiers.
- Pronouns – personal, reflexive, reciprocal, interrogative, relative, indefinite.
- Auxiliary Verbs – primary (be, have, do); modal (can, may, must…).
- Conjunctions – coordinating, correlative, subordinating.
- Complementizers – that, if, whether.
Phrasal Categories & Representation
- Phrasal categories (NP, VP, PP, AdjP, AdvP) are built recursively.
- Labelled-bracketing and tree diagrams visualise hierarchy.
Example sentence
(S
(NP (Det The) (Adj fat) (N cat))
(VP (V sat)
(PP (P on)
(NP (Det the) (N mat))))
)
Structural ambiguity: She attacked the man with a knife → two parses (instrument v. attribute).
Cross-Linguistic Insights & Flexibility
- Category universals vs language-specific realisation (e.g. grammatical gender in German vs natural gender in Sinhala).
- Syntactic flexibility – in Chinese a single form chi can be noun ‘food’ or verb ‘eat’; English needs derivation (run → runner).
- Implications – informs typology, L2 acquisition (article errors by Sinhala/Tamil speakers), NLP design.
Grammatical Categories
Number
- Core distinction {\text{singular},\;\text{plural}}.
- Some languages add dual (Arabic, Sanskrit) or trial.
Gender
- Grammatical gender (German: der Tisch MASC) vs natural/social gender (Sinhala lacks grammatical agreement).
Case (Classical 7-case model)
| Case | Core Role | Example (English gloss) |
|---|---|---|
| Nominative | subject | He came |
| Accusative | direct object | saw her |
| Dative | indirect object | gave a book to him |
| Genitive | possession | Maria's car |
| Instrumental | means | wrote with a pen |
| Locative | place | in the house |
| Vocative | address | John, come here |
| Modern English reduces to {\text{subject},\text{object},\text{possessive}}. |
Person
1st (person speaking), 2nd (addressed), 3rd (about). Drives agreement.
Voice
- Active – subject agent (The cat chased the mouse).
- Passive – subject patient (The mouse was chased by the cat). Alters information structure, allows topicalisation.
Tense & Aspect
- Tense locates event in time (past / present / future).
- Aspect views event as complete or ongoing.
- Perfective (completed) vs Imperfective / Progressive (ongoing).
- English auxiliaries: have (perfect), be + V-ing (progressive).
Verb-Phrase Combination Patterns
Four basic chains (max 4 auxiliaries):
- Modal + V ( can go )
- Perfect = have + V_{en} ( _has eaten_ )
- Progressive = be + V_{ing} ( _is eating_ )
- Passive = be + V_{en} ( _was eaten_ )
Combined: may have been being tested.
Worked Example: Decomposing a Sentence
Sentence: “The sergeant ordered the slave to be thrown into the water so that he could have experienced the true danger of life.”
- Morphemes – sergeant, order +-ed, slave, throw +-n, etc.
- Words – the | sergeant | ordered | the | slave | to | be | thrown | into | the | water | so that | he | could | have | experienced | the | true | danger | of | life.
- Phrases –
- NP: the sergeant, the slave, the water, the true danger.
- VP: ordered …, could have experienced ….
- PP: into the water.
- Clause: main + purpose clause (so that…).
Key Take-Aways
- Linguistic structure is hierarchical; mastery requires seeing the links across levels.
- Morphemes explain word formation; words combine into phrases whose heads determine category; clauses bring propositional force.
- Syntactic & grammatical categories (number, case, voice, etc.) regulate agreement and interpretation cross-linguistically.
- Representations (trees, brackets) reveal ambiguity and guide computational parsing.
- Comparative study and L2 research show both universals and language-specific flexibility, informing teaching, AI and theoretical models.
“When you know the parts and how they fit, you can build any sentence, in any language.”