Linguistics basics

In short

Linguistics is the systematic, scientific study of language — its sounds, word-forms, sentence structure, meaning and use — as opposed to the study of one language's literature or its prescriptive "correct" rules. It treats language as a rule-governed system that every native speaker knows unconsciously, and describes that knowledge at successive levels: phonetics, phonology, morphology, syntax, semantics, pragmatics and discourse. For a BA English student it is examined in the compulsory Language/Linguistics paper under UGC CBCS, in NTA UGC-NET English (Unit on Language and Linguistics), and in entrance tests for MA Linguistics and ELT programmes at EFL-U Hyderabad, JNU, Delhi University and the University of Hyderabad. In jobs it underpins content and localisation work, ELT and test-item writing, speech-and-language data annotation for Indian-language AI (Bhashini, AI4Bharat), subtitling and lexicography.

Explained

What linguistics is, and its core dichotomies

• Definition: the scientific study of human language — it observes data, forms hypotheses about rules, and tests them against native-speaker judgements.
• Descriptive vs prescriptive: descriptive linguistics records what speakers actually say ("Where you are going?" in Indian English); prescriptive grammar prescribes what they "should" say. Linguistics is descriptive.
• Langue vs parole (Ferdinand de Saussure, *Cours de linguistique générale*, 1916): *langue* is the shared abstract system of a speech community; *parole* is an individual's actual utterances.
• Competence vs performance (Noam Chomsky, *Aspects of the Theory of Syntax*, 1965): competence is the internalised grammar; performance is its use, with slips, hesitations and memory limits.
• Synchronic vs diachronic: synchronic describes a language at one point in time; diachronic traces its change across time.
• Syntagmatic vs paradigmatic relations: syntagmatic = how items combine in a chain (*the + tall + boy*); paradigmatic = how items substitute at one slot (*tall / short / clever*).
• Signifier and signified: Saussure's linguistic sign unites a sound-image (*signifier*) and a concept (*signified*); their link is arbitrary and conventional.
• Form vs function: the same form can do different jobs (*"Can you pass the salt?"* is interrogative in form, a request in function).

Properties of human language

• Arbitrariness: no natural resemblance between word and meaning (*dog*, *kutta*, *nāy*); onomatopoeia is a limited exception and even it is language-specific.
• Duality (double articulation): a small set of meaningless sounds combines into a large set of meaningful units.
• Productivity / creativity: finite rules generate infinitely many novel sentences.
• Displacement: language can refer to the absent, the past, the future and the fictional.
• Cultural transmission: the specific language is learnt from the community, not inherited genetically.
• Discreteness: sounds are categorically distinct — /p/ and /b/ differ in kind, not degree.
• Interchangeability, reflexiveness (language talking about language), broadcast transmission with directional reception — the fuller list is Charles Hockett's "design features".
• Universals: all languages have vowels and consonants, nouns and verbs, negation, questions and recursion; Universal Grammar is Chomsky's claim that these follow from an innate faculty.

Levels of linguistic analysis

1. Phonetics — the physical production, transmission and perception of speech sounds; articulatory, acoustic and auditory. (Detailed treatment belongs to the separate phonetics note.)
2. Phonology — the sound *system* of a particular language: which contrasts are meaningful, and how sounds pattern.
   • Phoneme: the smallest unit that distinguishes meaning — established by minimal pairs (*pin/bin* → /p/ vs /b/).
   • Allophone: a positional variant of one phoneme that never changes meaning — English /p/ is aspirated [pʰ] initially (*pin*) and unaspirated [p] after /s/ (*spin*). In Hindi, aspiration is phonemic: पल /pal/ vs फल /pʰal/ are different words — the classic exam contrast.
   • Complementary distribution (allophones), free variation, phonotactics (permitted sequences: English allows /str-/, not /ŋ-/ initially).
   • Suprasegmentals: stress, tone, length, intonation; English is stress-accented, Punjabi and Mizo are tonal.
3. Morphology — internal structure of words (see next sub-heading).
4. Syntax — arrangement of words into phrases and sentences.
5. Semantics — literal, context-free meaning of words and sentences.
6. Pragmatics — meaning in context: speaker intention, implicature, speech acts, deixis.
7. Discourse analysis — structure and cohesion of texts and conversations above the sentence.
8. Mnemonic order of "size": sound → word-part → word → phrase → sentence → utterance → text.

Morphology

• Morpheme: the smallest meaningful unit. *unhappiness* = un- + happy + -ness (three morphemes). A morpheme is not a syllable: *elephant* has three syllables, one morpheme.
• Free vs bound: free morphemes stand alone (*book*, *run*); bound cannot (*-ed*, *re-*, *-s*).
• Root, stem, base, affix: affixes are prefix (*un-*), suffix (*-ness*), infix (rare in English; *-um-* in Tagalog), circumfix.
• Inflection vs derivation:
  • Inflection marks grammatical categories without changing word class or core meaning: *book → books*, *walk → walked*. English has only eight inflections (*-s* plural, *-'s* possessive, *-s* 3sg, *-ed* past, *-ed/-en* participle, *-ing*, *-er*, *-est*).
  • Derivation makes a new word, often of a new class: *nation → national → nationalise → nationalisation*.
  • Order: derivation is inside, inflection outside — *nationalisations*, never **nationsalisation*.
• Allomorph: variants of one morpheme — English plural /-s/ in *cats*, /-z/ in *dogs*, /-ɪz/ in *boxes*; past *-ed* as /t/, /d/, /ɪd/. Zero allomorph in *sheep → sheep*; suppletion in *go → went*.
• Word-formation processes: compounding (*blackboard*), blending (*smog*, *Hinglish*), clipping (*ad*, *flu*), acronym/initialism (*UPSC*, *NEFT*), back-formation (*edit* from *editor*), conversion/zero-derivation (*to google*), reduplication (very common in Indian languages: Hindi *garam-garam*, *chai-pani*), borrowing (*jungle*, *shampoo*, *bungalow* into English from Indian languages).
• Morphological typology: isolating/analytic (Mandarin, little inflection), agglutinating (Tamil, Kannada, Turkish — long strings of clear-cut suffixes), inflecting/fusional (Sanskrit, Latin — fused endings), polysynthetic (Inuktitut). Tamil *vandhaargal* (came-HON.PL) is a standard agglutination example.

Syntax

• Constituency: sentences are built of nested phrases, not flat word-strings; tested by substitution, movement and standing alone as an answer.
• Phrase types: NP, VP, AdjP, AdvP, PP — each headed by its own category.
• Immediate Constituent (IC) analysis and phrase-structure rules:
  • S → NP VP; NP → (Det) (Adj) N (PP); VP → V (NP) (PP); PP → P NP.
• Tree diagrams: draw labelled nodes; every branching node must have a label and a head.
• Deep vs surface structure: the active *The police arrested the thief* and passive *The thief was arrested by the police* share a deep structure; transformations relate them (Chomsky's Standard Theory, 1965).
• Ambiguity: lexical (*bank*), structural (*old men and women*; *I saw the man with the telescope*) — structural ambiguity is shown by drawing two different trees.
• Grammaticality vs meaningfulness: *Colourless green ideas sleep furiously* is grammatical but semantically odd; **Sleep ideas green furiously colourless* is ungrammatical (marked with *).
• Word order typology: English is SVO; Hindi, Bengali, Tamil, Kannada and most Indian languages are SOV with postpositions, verb-final clauses and head-final relative clauses.
• Later frameworks examined in outline: Government and Binding (1981), X-bar theory, Principles and Parameters, Minimalist Program (1995); functional alternatives — Halliday's Systemic Functional Grammar (language as choice: ideational, interpersonal, textual metafunctions), Tagmemics, Case Grammar (Fillmore).

Semantics and pragmatics

• Sense vs reference: *sense* is the place of a word in the system; *reference* is the real-world entity. *The President of India* has constant sense, changing reference.
• Denotation vs connotation: *politician* denotes an office-holder, connotes (often) opportunism.
• Sense relations: synonymy (*big/large* — rarely total), antonymy (gradable *hot/cold*; complementary *alive/dead*; relational *buy/sell*), hyponymy (*rose* is a hyponym of *flower*, which is the hypernym), meronymy (part–whole: *wheel–car*), homonymy (same form, unrelated senses: *bank*), polysemy (one word, related senses: *head* of body/company/river), homophones (*sea/see*), homographs (*lead*).
• Semantic features / componential analysis: *man* = [+human, +adult, +male]; *girl* = [+human, −adult, −male].
• Semantic change: widening (*holiday* from holy day), narrowing (*meat* once any food), amelioration (*knight*), pejoration (*villain*), metaphor, metonymy.
• Sentence semantics: entailment (*He was murdered* entails *He died*), presupposition (*Have you stopped smoking?* presupposes earlier smoking), contradiction, tautology, anomaly.
• Pragmatics: speech acts (J. L. Austin's locutionary / illocutionary / perlocutionary; Searle's five classes — representatives/assertives, directives, commissives, expressives, declarations); performatives and felicity conditions; deixis (person, place, time, discourse, social — Hindi *tū/tum/āp* is social deixis).
• Grice's Cooperative Principle and four maxims — Quantity, Quality, Relation (relevance), Manner; flouting a maxim creates implicature (*"The food was warm"* implies it should have been hot).
• Politeness: Brown and Levinson's positive and negative face; Indian English prefers indirect requests (*"Kindly do the needful"*).
• Discourse: cohesion (reference, substitution, ellipsis, conjunction, lexical cohesion — Halliday and Hasan) vs coherence; turn-taking and adjacency pairs.

Historical, comparative and Indian sociolinguistics

• Language change operates at every level; comparative method reconstructs proto-forms from systematic correspondences.
• Indo-European and Sanskrit: Sir William Jones's 1786 address to the Asiatic Society, Calcutta, noted Sanskrit–Greek–Latin kinship and founded comparative philology. Grimm's Law (Germanic consonant shift: PIE *p t k* → *f θ h*; Sanskrit *pitṛ* ~ English *father*), Grassmann's Law and Verner's Law are the standard examples.
• Indian grammatical tradition: Pāṇini's *Aṣṭādhyāyī* (c. 5th–4th century BCE), about 4,000 sūtras — the earliest generative, rule-ordered grammar; Bhartṛhari's *Vākyapadīya* (sphoṭa theory); Tolkāppiyam for Tamil.
• Language families of India (four major): Indo-Aryan (Hindi, Bengali, Marathi, Gujarati, Punjabi, Odia, Assamese — about three-quarters of speakers), Dravidian (Tamil, Telugu, Kannada, Malayalam), Austro-Asiatic/Munda (Santali, Mundari, Khasi), Tibeto-Burman/Sino-Tibetan (Manipuri/Meiteilon, Bodo, Mizo); the Andamanese languages and Tai-Kadai (e.g. Khamti) are treated as further, much smaller families.
• Constitutional and policy position: Article 343 — Hindi in Devanagari as the Union's official language, with English continued for official purposes under the Official Languages Act, 1963; Articles 345–347 for State languages; Article 350A — mother-tongue instruction at primary stage; Article 351 — development of Hindi. The Eighth Schedule currently lists 22 languages (Sindhi 1967; Konkani, Manipuri, Nepali 1992; Bodo, Dogri, Maithili, Santali 2003) — further additions are periodically proposed, so verify the count against the latest amendment. Classical Language status has been conferred on Tamil, Sanskrit, Telugu, Kannada, Malayalam, Odia and, in October 2024, Marathi, Pali, Prakrit, Assamese and Bengali — eleven in all, subject to further Cabinet decisions.
• NEP 2020 mandates a three-language formula with mother tongue / regional language as medium where possible, and set up the Bharatiya Bhasha Samiti; the Three-Language Formula itself dates to 1968.
• Sociolinguistic concepts: dialect vs language (mutual intelligibility and the dialect continuum — Hindi–Urdu as one *Hindustani* continuum with two registers and scripts), idiolect, sociolect, register (field, mode, tenor), standard language and codification, isogloss.
• Bilingualism and multilingualism: diglossia (Ferguson — High/Low varieties; Tamil's literary *centamiḻ* vs spoken *koṭuntamiḻ*; Arabic), code-switching (inter-sentential) vs code-mixing (intra-sentential — "Hinglish"), borrowing, pidgin (reduced contact language, no native speakers) → creole (nativised, fully expanded; Nagamese is an Indian example often cited as a pidgin/creole), lingua franca (English and Hindi in India), language shift, endangerment and death — UNESCO and the PLSI (People's Linguistic Survey of India, 2010s) record about 780 Indian languages, many with very small speaker bases; Census 2011 lists 121 languages with 10,000+ speakers and 1,369 rationalised mother tongues.
• Variation and attitudes: Labov's quantitative sociolinguistics; Bernstein's restricted/elaborated codes; Sapir–Whorf hypothesis (strong determinism rejected, weak relativity debated); Kachru's Three Circles — Inner (UK, USA), Outer (India, Nigeria), Expanding (China, Japan) — placing Indian English as an institutionalised second-language variety with its own features (retroflex stops, syllable-timed rhythm, *isn't it?* as invariant tag, *prepone*, *batchmate*).

Applied and interdisciplinary branches

• Applied linguistics / ELT: syllabus design, materials writing, error analysis, contrastive analysis, Interlanguage (Selinker), Krashen's input hypothesis, CLT and task-based learning; CEFR levels A1–C2 referenced by Indian test providers.
• Psycholinguistics: language acquisition stages (cooing, babbling, one-word/holophrastic, two-word, telegraphic), over-generalisation (*goed*), critical period, aphasias — Broca's (non-fluent, agrammatic) vs Wernicke's (fluent, meaningless), lateralisation to the left hemisphere.
• Neurolinguistics, clinical linguistics and speech–language pathology (RCI-registered practice in India).
• Computational linguistics and NLP: tokenisation, POS tagging, morphological analysers (critical for agglutinating Indian languages), parsing, machine translation, ASR and TTS; Indian resources — AI4Bharat (IIT Madras) IndicNLP corpora and IndicTrans, Bhashini/National Language Translation Mission, CIIL Mysuru corpora, TDIL programme of MeitY, Anusaaraka and Sampark MT systems.
• Forensic linguistics (authorship attribution, disputed statements), stylistics, lexicography (dictionary-making, corpus-based), translation studies, language documentation.

Key points

• Linguistics — scientific, descriptive study of language structure and use, not prescription of correct usage.
• Langue / parole — Saussure's system vs actual speech; competence / performance is Chomsky's cognitive restatement.
• Arbitrariness — no inherent link between signifier and signified; the sign is conventional.
• Duality — meaningless sounds combine into meaningful units, giving unlimited output from limited means.
• Phoneme vs allophone — meaning-distinguishing unit vs its context-bound variant; proved by minimal pairs.
• Minimal pair — two words differing in exactly one sound in the same position (*pin/bin*).
• Morpheme — smallest meaningful unit; allomorph — its phonetic variants (*/s/, /z/, /ɪz/* plural).
• Inflection vs derivation — grammatical marking with no class change vs new word, often new class; derivation precedes inflection.
• Phrase-structure rules and trees — S → NP VP etc.; structural ambiguity equals two possible trees.
• Deep vs surface structure — one underlying meaning, several surface forms, linked by transformations.
• SOV vs SVO — Indian languages are typically SOV with postpositions; English is SVO with prepositions.
• Typology — isolating, agglutinating (Dravidian), fusional (Sanskrit), polysynthetic.
• Sense relations — synonymy, antonymy, hyponymy, meronymy, homonymy, polysemy.
• Entailment vs presupposition vs implicature — logical consequence, assumed background, and inference from flouting Grice's maxims.
• Speech acts — locutionary, illocutionary, perlocutionary (Austin); Searle's five illocutionary classes.
• Diglossia, code-switching, pidgin/creole — core contact phenomena; code-mixing is intra-sentential, switching inter-sentential.
• Indian framework — four major families; Eighth Schedule 22 languages; Articles 343–351; NEP 2020 three-language formula.
• Pāṇini's Aṣṭādhyāyī — the earliest formal generative grammar, with rule ordering and a metalanguage.

Examples

Example 1 — Morpheme and allomorph analysis. Analyse *internationalisations*.

1. Root: *nation* (free morpheme).
2. *nation* + -al → *national* (noun → adjective, derivational).
3. *inter-* + *national* → *international* (prefix, derivational).
4. + -ise → *internationalise* (adjective → verb, derivational).
5. + -ation → *internationalisation* (verb → noun, derivational).
6. + -s → plural (inflectional, outermost).

Answer: six morphemes — one free root and five bound affixes, of which four are derivational and the final *-s* inflectional; the order confirms that inflection attaches outside derivation. Allomorphy: the plural here is /z/ (after the voiced /n/); in *contexts* it is /s/, in *buses* /ɪz/ — three allomorphs of one plural morpheme.

Example 2 — Phoneme vs allophone, English and Hindi. Data: English [pʰɪn] *pin*, [spɪn] *spin*; Hindi [pal] 'moment', [pʰal] 'fruit'.

1. In English, [pʰ] and [p] never occur in the same environment — [pʰ] word-initially before a stressed vowel, [p] after /s/. This is complementary distribution, and no pair of English words is distinguished by aspiration alone.
2. Conclusion: [pʰ] and [p] are allophones of one phoneme /p/ in English.
3. In Hindi, [pal] and [pʰal] form a minimal pair — the sole difference is aspiration, and the meaning changes.
4. Conclusion: /p/ and /pʰ/ are separate phonemes in Hindi.

Takeaway: phonemic status is language-specific; the same physical sounds can be contrastive in one language and variant in another. (Practical use: Hindi speakers hear English aspiration as meaningful, which explains predictable L1-transfer errors in ELT materials.)

Example 3 — Structural ambiguity resolved by trees. Sentence: *The teacher taught the students with charts.*

1. Reading A — PP modifies the VP (instrument): the teacher used charts to teach. S → NP[The teacher] VP[V[taught] NP[the students] PP[with charts]]
2. Reading B — PP modifies the NP (attribute): the students who had charts were taught. S → NP[The teacher] VP[V[taught] NP[NP[the students] PP[with charts]]]
3. Test: substitute a pronoun for the object. *The teacher taught them with charts* keeps only Reading A; *The teacher taught them* where *them* = "the students with charts" keeps Reading B — pronoun substitution isolates the constituent.

Answer: two distinct trees, hence genuine structural (not lexical) ambiguity — the words and their senses are unchanged.

Example 4 — Implicature and maxims. A reference letter reads, in full: *"Mr Rao attended my classes regularly and his handwriting is excellent."*

1. Literally (semantics) nothing negative is stated; both propositions may be true.
2. The maxim of Quantity requires as much information as is needed; a reference should address ability.
3. By under-informing where fuller information was expected, the writer flouts Quantity.
4. The hearer assumes the Cooperative Principle still holds and infers: the writer has nothing good to say about the candidate's ability.

Answer: a conversational implicature of a negative assessment — cancellable (the writer could add "and he is also the best student I have taught"), unlike an entailment, which cannot be cancelled.

Example 5 — Comparison tables.

Where it is used

• Language AI and data annotation: AI4Bharat (IIT Madras), Bhashini/NLTM, Reverie, Gnani.ai, Sarvam AI and Google/Microsoft India hire linguists for POS/NER tagging, transcription guidelines, speech corpora and MT evaluation across the Eighth Schedule languages.
• ELT, test development and assessment: item writing and rubric design for CBSE, NTA, IELTS/TOEFL centres, British Council and Cambridge partners; EFL-U Hyderabad and CIEFL-lineage institutions train for this.
• Publishing, lexicography and localisation: Oxford University Press India, Orient BlackSwan, dictionary and corpus projects, subtitling and dubbing for OTT platforms, UI localisation for Indian-language apps and government portals.
• Media, content and UX writing: style guides, editorial standards, voice-assistant and chatbot script design, where register and politeness in Indian English matter commercially.
• Public sector and policy: CIIL Mysuru, Census of India language division, Sahitya Akademi, NCERT textbook and language-policy work under NEP 2020, Linguistic Data Consortium for Indian Languages (LDC-IL).
• Clinical and forensic work: speech–language pathology in RCI-recognised set-ups (AIISH Mysuru), and forensic authorship analysis for legal and investigative agencies.

Interview questions

• Q: How does linguistics differ from grammar as taught in school? A: School grammar is prescriptive and tells you which forms are correct; linguistics is descriptive and explains the system underlying what speakers actually produce, including non-standard varieties.
• Q: What is the difference between a phoneme and an allophone? A: A phoneme is a contrastive sound unit that distinguishes meaning, shown by minimal pairs; an allophone is a predictable variant of a phoneme in a particular environment and never changes meaning.
• Q: Give an Indian example where the phoneme inventory differs between two languages. A: Aspiration is phonemic in Hindi (*pal* 'moment' vs *phal* 'fruit') but only allophonic in English, where [pʰ] and [p] are variants of /p/.
• Q: What is the difference between competence and performance? A: Competence is a speaker's internalised knowledge of the rules of the language; performance is the actual use of that knowledge, affected by memory, attention, fatigue and slips.
• Q: Is a morpheme the same as a syllable? A: No — *elephant* has three syllables but one morpheme, while *cats* has one syllable and two morphemes; syllables are phonological units, morphemes are meaningful ones.
• Q: How would you show that a sentence is structurally ambiguous? A: By drawing two different phrase-structure trees in which the same words group differently — for example whether a prepositional phrase attaches to the verb phrase or to the object noun phrase.
• Q: What is the difference between entailment and implicature? A: Entailment follows logically from the sentence and cannot be denied without contradiction; implicature is inferred from context and the cooperative principle and can be cancelled by the speaker.
• Q: Explain diglossia with an Indian example. A: Diglossia is the use of two varieties of one language in complementary domains — in Tamil, the literary variety is used in writing, news and formal speech, and the colloquial variety at home and among friends.
• Q: Is Indian English a deficient variety of English? A: No — it is an institutionalised Outer-Circle variety in Kachru's model, with its own stable phonological, lexical and syntactic norms, acceptable in its own context while differing from British or American standards.
• Q: What is the difference between a pidgin and a creole? A: A pidgin is a simplified contact language with no native speakers, used for limited purposes; a creole is a pidgin that has acquired native speakers and expanded into a full language.
• Q: Why is morphology harder for Indian-language NLP than for English? A: Dravidian and many Indo-Aryan languages are agglutinating, so one word may carry several suffixes for case, number, person, tense and honorific — requiring a morphological analyser rather than simple whitespace tokenisation.
• Q: Why is Pāṇini's grammar important to modern linguistics? A: The *Aṣṭādhyāyī* describes Sanskrit through about 4,000 ordered, economical rules with a formal metalanguage — effectively a generative grammar, which later influenced formal grammar and computational approaches.
• Q: What does the arbitrariness of the sign mean, and where does it break down? A: There is no natural connection between a word's form and its meaning, which is why languages use different words for the same concept; onomatopoeia and sound symbolism are partial exceptions, and even they are shaped by each language's sound system.

Common mistakes

• Treating linguistics as prescriptive grammar — writing answers about "correct usage" when the question asks for a description of a system.
• Confusing phoneme with letter, and allophone with phoneme — then claiming aspiration is phonemic in English.
• Counting syllables as morphemes, or missing the inner derivational layers in long words like *internationalisations*.
• Reversing the order of affixation — placing inflection inside derivation, or asserting English has many inflections (it has eight).
• Mixing up entailment, presupposition and implicature, or treating implicature as part of literal semantics.
• Drawing trees without labelling nodes, or without showing the second tree when asked to prove structural ambiguity.
• Calling Hindi–Urdu, or a regional variety, a "dialect" in a pejorative sense; the dialect–language distinction is sociopolitical, not purely linguistic.
• Quoting outdated Indian figures — the Eighth Schedule count and the list of Classical Languages have both changed, so state them as "currently" and note that additions are made by amendment or Cabinet decision.

Related topics

Phonetics and spoken English · Grammar and sentence correction · History of the English language · Indian writing in English · Stylistics and literary theory · Computational linguistics and NLP for Indian languages