Linguistics - Finals

Syntax 1 

1.Basic Properties of Syntax

Syntax- pertaining to sentences 

• Syntactic knowledge is highly productive (evidence: nearly every sentence we produce is brand new) 

• Therefore we must have a syntactic grammar and know syntactic rules 

Properties of syntactic rules

• Recursive

• Based on syntactic categories

• Hierarchical

Recursion

• When something is defined in terms of itself 

• When a syntactic structure can contain itself (They will major in psychology = I think [they will major in psychology])

Example 1: “I think” is a sentence, and “they will major in psychology” is also a sentence. “I think they will major in psychology” are two sentences that are within each other

• Sentences can contain sentences 

Example 2: “The mother of my brother” is another example of recursion 

• ‘My brother’ is a noun phrase, and ‘the mother of’ is also a noun phrase. Therefore, “the mother of my brother” has a noun phrase that contains a noun phrase

• All/almost all languages have recursion

Infinite productivity

• Due to recursion, there are an infinite number of possible grammatical sentences in a language 

  • You can always add “I think that” to a sentence, turning it into a new sentence 

  • Afterwards, you can add “he knows I think that” to make it an even longer sentence

Syntactic categories

Syntactic knowledge - relates to parts of speech

• We don’t memorise what others words a specific word can connect to, because language is productive and every sentence that we say has virtually never been said before

• Therefore we can always connect a word with a different word (even though it has never been done before) 

  • Therefore therefore, we cannot use word-to-word knowledge because that is not productive enough

• Instead, our syntactic knowledge is related to parts of speech and not individual words 

  • Example: Adjective → adjective → noun (big green fox) 

Syntactic categories - Noun, adjective, adverb, verb, noun phrase… etc. 

Syntactic knowledge is hierarchical 

• Sentences contain long-distance dependencies (if…then)

• Collections of words behave as a single functional unit 

  • Either [the girl eats ice cream] or [she eats candy] 

Terminology

Expression/utterance - one or more words, not necessarily grammatical

Sentence - a string of words that is grammatical in a language

Subject - the expression that comes before a verb

Object - the expression that comes after a verb

2.Phrase Structure Grammars

• Hierarchical grammars that utilise syntactic categories

  • Details the syntactic rules of a language

  • Explains how phrase structures are built

  • Explicitly and specifically describes the mental syntactic knowledge of the language’s speakers

X → YZ 

• NP → Pronoun 

• NP → {pronoun, proper noun, determiner noun) 

• Triangles are a notational device to indicate that we’re not bothering with drawing the internal structure 

Overview

S = sentence → NP VP

NP → Proper noun, pronoun, mass noun, determiner noun, 

N → adjective noun, noun + preposition phrase

VP → prepositional phrase 

• All sentences consist of NP VP

• Nouns are things that must be preceded by a determiner 

Verbs

• Have different arguments (what is required to also be present for the sentence to be grammatical

ITV (intransitive) → no complement (nothing follows suit) 

• Sleep, dream, look

TV (transitive) → NP follows

• Devour (I will devour this meal, where “this meal” is a NP)

• Want (something), wash (something, push (something) 

DTV (ditransitive) → NP NP follows

• Ask (I ask you to get me that book, where ‘get me’ is an NP, and ‘that book” is the second NP) 

• Gave [x] the [y]

• Ask [x] the [y]

SV (sentential) → a sentence follows (NP VP) 

• I believe she will run tomorrow (‘she will run tomorrow’ is a sentence that can stand on its own) 

• Think, believe, hope

Mass nouns

• Water, sand, furniture 

• Usually can’t be counted (can be preceded by ‘some’ but not ‘a’) 

  • Some water, some sand, some furniture 

• Does not take the plural suffix (some exceptions apply) 

• Often are substances 

Count nouns

• Water bottle, chair 

• Can be preceded by ‘a’ and number words 

  • A water bottle, two chairs 

• Takes the plural suffix when plural (the chairs, the water bottles) 

• Usually things that can be counted 

Parsing a Sentence

• When creating a tree for a sentence, we are: 

  • Determining which syntactic rules of the language apply to produce the sentence 

  • Explicitly indicating constituent structure and modification relations 

How to parse a sentence

• Think like a computer – follow the rules in the grammar 

  • Don’t use your intuition! 

• Each rule (and the treelets it generates) must fit together like a puzzle from start to finish (S to terminal nodes) 

1. Label the part of speech of each word in the sentence (wait on the verb) 

2. Place S → NP VP at the top of your tree 

3. Everything before the verb will be in the NP, and the verb and everything to the right will be in the VP

4. Build the NP 

5. Build the VP

   1. Determine the verb subcategorisation frame (ITV, TV, DTV, S)

   2. Some verbs can behave in more than one way, so use the part of speech relevant to the sentence you’re analysing

Example: I asked [her] a [question] = ditransitive

• I asked [her] = transitive

• I asked = intransitive

Syntax 2 - Contemporary Theory

Phrase Structure Grammars

• Can be recursive 

• Refer to syntactic categories

• Are hierarchical

With a PSG as a theory of our syntactic grammar, a sentence is constructed as follows: 

1. Syntactic rules are used to create a syntactic tree

2. Words are inserted into the tree 

Issues with PSGs as a theory of syntactic grammar

• While they have the right syntactic properties, they are not well-integrated with cognition 

• How are sentences actually constructed? 

  • Starting with rules - how do the rules know what sentence we want to express? 

• How do we build the right tree? 

  • How do we know the subcategorisation frame of a verb ahead of time? 

New Approach

• Instead of just things to be inserted into a tree, words are in fact small pieces of syntactic structure 

• Instead of starting with rules and working down to words, start with words and work your way up to structure (bottom up method) 

Sketch

1. Lexical items are selected based on how well they fit the meaning to be expressed 

2. Lexical items project syntactic structure (elementary trees) 

3. To build the sentence, merge the treelets appropriately 

Projection - syntactic structure is stored and retrieved for each word 

Pros

1. Simplicity in mechanism: project and merge 

2. More explicit means for lexical terms to influence syntactic structure 

Now that lexical terms are now elementary bits of syntactic structure, what does that structure look like? 

• Does every word have a different syntactic structure? 

• Does every word have the same type of syntactic structure? 

X-bar theory

• Every phrase in every sentence in every language is organised the same way

• Syntacticians noticed that all phrases seem to need to have an intermediate level of structure between the word [x] and the phrase [xp] 

  • X’ intermediate level (this is x-bar)

Theory

• Phrases are the building blocks of sentences 

Headedness principle: every phrase has a head, which determines its syntactic category

• X is a variable replaced with the part of speech of the head 

• Every phrase has an intermediate structure called X’

• Every phrase may contain other phrases in specifier and complement position 

X-bar theory, what are sentences?

• In PSG, sentences are the root node (S → NP VP) 

• In X-bar, every syntactic structure is a phrase, and every phrase has a head

  • The defining characteristic of a sentence is inflection (in English, it’s tense)

Sentences are inflectional phrases (IP)

• I → tense feature 

• Specifier → subject NP

• Complement → VP

Sentences are also wrapped in a complementiser phrase 

• CP - phrase that houses sentences, specifies whether the sentence is declarative, interrogative, imperative, or subjunctive 

  • Also help relate embedded sentences to the main sentence

  • Example: that, if, (I think that… I wonder if…)

Summary

• In x-bar theory, sentences are constructed through Project and Merge

  • Words project syntactic structure (make treelets)

  • Syntactic structure is merged together 

• Syntactic structure has X-bar structure 

  • Every phrase in every language is organised the same way 

  • Basic unit of syntax is a phrase 

  • Every phrase has a head (called x’ level) 

Example: I [head, present] think that she [head, past] deserved a vacation

• Where ‘I think that’ is a phrase, and ‘she deserved a vacation’ is another phrase 

• The head of a sentence is inflection (tense) 

  • All sentences are inflection phrases (IP)

• All sentences are embedded in a complementiser phrase (CP)

Syntax 3 - Contemporary Theory

1. Changing how sentences are constructed

Phrase structure grammars 

• Have the right properties for syntactic knowledge (hierarchical, lexical categories recursive) 

• However, they are not especially well-integrated with cognition 

  • How are sentences actually constructed (S → NP VP)

  • Mechanisms may be unclear (subcategorisation frames are rules?)

Movement

Our theory so far consists of building a sentence using project and merge 

• However, there are inflectional phrases, complementiser phrases, and movement to consider 

Terminology

Auxillary verbs - verbs and modals that come before the main verb 

Example: should, would, will, did

• Changes declarative sentences into yes/no questions (Trevante will give Jamila the ball = Will Trevante give Jamila the ball?)

• This is subject-auxiliary inversion, where the subject and auxiliary verb swap positions 

Wh- questions

• Same thing as subject-auxiliary inversion, but a wh-word will be at the beginning of the question 

• The dog will eat bacon = What will the dog eat? 

This shows that different sentence types have very similar syntax

Data: Sentence position and role

• There is usually a very simple relationship between sentence position and thematic role

Subject: entity doing the action (Agent) 

Object: entity undergoing the action (patient/experiencer) 

Example 1: To please

The teacher pleases his students 

• Subject - the teacher = agent 

• Object - his students = experiencer (they experience the pleasure) 

Example 2: To push

They pushed the heavy sled 

• Subject - they = agent 

• Object - sled = patient (sled is being pushed) 

Summary

• Different sentence types ultimately aren’t all that different (subject-auxiliary switch to make a yes/no sentence) 

• There is sometimes a misalignment between position and thematic role 

• Wh- words have the same role as the things they replace

Movement

• Can explain the similarly in structure and how elements can systematically receive their thematic roles if we assume that we 

1. Build a basic type of declarative sentence 

2. Assign thematic roles based on these positions

3. Minimally move elements to form other sentence types (ex. questions) 

How does movement explain the similarity between declarative and question sentences

• Declarative sentences are built by project and merge and nothing else happens 

• Question sentences are built by 

  • Generating the basic declarative structure using project and merge

  • Moving the auxiliary/wh-word to the complementiser phrase 

  • The auxiliary moved to the C, and the wh-word moves to the specifier of C

Explaining wh-questions

• Wh-words are generated where they would be in a declarative sentence and moved to the specifier of C

The dog will eat food = What will the dog eat?

1. Project and merge generate the sentence with ‘what’ in the second NP position 

2. To form the question, movement rules move ‘what’ to the specifier of C position 

• Since the ‘will’ is in the auxiliary position, it also moves to C to form the question, just like in subject auxiliary inversion

• With this approach, the similarity between declarative sentences and question sentences is explained 

  • They are similar since they are generated with the exact same structure and then questions are made by moving only 1-2 words

• We form questions by raising the auxiliary to C because CP (complementiser phrases) determine the sentence type

  • By moving the auxiliary to C (the head of the CP), it can now dictate that it is a question-type sentence

How are thematic roles systematically assigned?

• In a theory without movement, it’s not easy to explain why most of the time subjects are agents and objects are patients/experiences but sometimes not 

• Also not easy to explain why wh-words have the same thematic role as the words they are ‘standing in’ for 

Thematic role assignment

In a theory of movement, it makes things simple and we don’t have to use idiosyncratic assignment rules 

1. Thematic roles are assigned in specific places in the syntactic structure 

• Agent is assigned in the subject NP position

• Experiencer/patient is assigned in the object NP position 

2. Words are generated in the correct position to receive the correct thematic role

3. Movement may then move them to a different position

• In this way, instead of having all of the words in a proper sentence, the words are generated in their own place where they have roles (there is a place for agents, experiencers…) 

  • They are moved to then make a sentence

Movement helps explain

• Why different sentences types are so similar 

• How to think about elements receiving their thematic roles no matter their position 

Subcategorisation frames

Some verbs are obligatorily transitive (they need an object NP)

• I will devour the cake, I will award the prize. 

• ‘The cake” and ‘the prize” are the object NP 

Except in wh-questions they don’t seem to have an object NP

• What will I devour? → I will devour ‘the cake’ (which is replaced with a wh-question)

• What will I award?

• In a theory without movement, every sentence is generated through project and merge 

• We would need one set of subcategorisation frames for declarative sentences and another for wh-question sentences

• This is too complicated, so the theory of movement makes more sense than not having the theory

• This misses the fact that the number of objects in wh-questions is very specific 

  • They always have 1 less than declaratives 

Example 1: I will award [the prize] = What will I award? 

Example 2: Trevante will give [Jamila] [the ball] = What will Trevante give [Jamila]?

Explanatory power

This fact is not predicted by a theory without movement, but it is predicted by a theory of movement 

• Verbs always have the right number of objects

• In wh-questions, one of them has the form of a wh-word and moved out of its original object position

• We only need to have 1 subcategorisation frame for each verb (any sentence that has a verb that isn’t an intransitive verb can be turned into a wh-question) 

  • Movement explains why we can’t make wh-questions out of intransitive verbs 

You smiled ≠ what did you smile? 

• Smile is an intransitive verb, so nothing is in the object position after that. Therefore it cannot be turned into a wh-question. No place for the wh-word to be generated in and moved from

Blocking

• If movement is real (actually a part of how sentences are built) it should only happen if the destination location is empty 

• Movement should be blocked if a word already happens to be in the destination location

• According to our theory, all sentences are enclosed in a CP. Yes/no questions are formed by moving the auxiliary from I to C 

• C may be filled by complementiser ‘that’ when there is an embedded sentence

The theory of movement explains why “should coke be cheaper” is grammatical, but “she that that should coke be cheaper” is ungrammatical 

• “Should” has a landing site in the basic sentence (an empty C) but does not in the embedded sentence (C is filled by ‘that’), blocking it from moving 

Summary

• Although theoretical elements like Inflectional Phrases (IP), Complementiser Phrases (CP), and Movement may at first seem unmotivated and unnecessarily complicated 

  • Their existence can be motivated from real data (complimenters in Swahili and other languages)

  • They can have powerful explanatory value

• Pattern for X- bar = N, N’, NP (V, V’, VP… and so on) 

• Movement talks about tense, and x-bar is about making treelets

Lexical Semantics

Semantics - subfield of linguistics that studies meaning 

• Meaning = concepts and semantics 

Concepts: 

• Non-linguistic meaning

• Animals and pre-lingual infants have concepts 

• The conceptual system

Semantics:

• Meaning expressed by linguistic utterances 

• Semantics helps us carve up this information up into the space that language describes

Branches of semantics

• Lexical semantics: focus on words (how they have meaning)

• Compositional semantics: focus on how words combine (to make meaning)

Semantics: Focus on study

• Semantics focuses on the ‘ordinary’ meaning of words and phrases 

• “Literal” meaning

• Depends on lexical meaning and syntax

• Semantics focuses on describing how humans expressing their understanding of the world by means of linguistic utterances 

Lexical Semantics

1. Sense and reference

2. Semantic relations among words

3. Theories of lexical meaning

Lexicalisation

• Semantics carves up meaning for language 

• Concepts are presumably universal, and semantics is language-specific 

• This plays out in differences in lexicalisation (how concepts map to words)

• If a language has a word for a particular concept, we say that the concept has been lexicalised in that language 

• The same concept may be lexicalised in one language and not lexicalised in another 

Example: In Spanish, there is a word for the concept “to wear something for the first time”, but English, the concept is there but it isn’t lexicalised

• Do people who have words for concepts think about things differently compared to others who don’t have words for concepts in their language 

• This is called linguistic determinism

  • No solid evidence that people with different language and different lexicalisation don’t really think that differently

Example 1: In english, we just use at to say that we’re somewhere. In Polish, there are four different variations for ‘at’ depending on whether it’s spatial/locational, functional, temporal, or other

Section 2: Semantic relations

Synonyms - words that have the same meaning 

Example: purchase/buy, big/large, remember/recall

Antonyms - words that have the opposite meaning

Example: dark/light, hot/cold, arrive/depart 

Types of antonyms

Complementary: referents must be either picked out by one expression or the other (not both) 

Example: married/single, visible/invisible

Gradable: represent points on a continuum (something may be neither one nor the other) 

Example: warm/cold, smal/big

Reverses: one ‘undoes’ the other 

Example: expand/contract, ascend/descend 

Converses: for one to occur, the other must occur as well

Example: lend/borrow, employer/employee

Semantic relations

Hypernyms: a word whose meaning includes the meaning of a subordinate concept (larger class) 

Example: furniture is a hypernym of chair, utensils is a hypernym of fork

Hyponym: a word whose meaning is included or is a subset of the meaning of another concept (a specific instance) 

Example: chair is a hyponym of furniture 

Meronym: a word that is a part of another word 

Example: mast is a meronym of sailboat

Holonym: a word that contains another word (there is a mast inside a sailboat)

Example: sailboat is a holonym of mast 

Words with multiple meanings

Polysemy - a word that has several related meanings

Example: Bright = shining/intelligent, paper = article or news/material for writing, chicken = the animal/coward

Homophony - a word that has several unrelated meanings

Example: light = not heavy/form of illumination, bank = financial institution/side of a river, club = social organisation/weapon

Section 3: Sense and Reference

Reference

• One of the things that words do is refer to things in the world 

  • Reference/denotation are the things in the world it refers to 

  • Referent is a specific thing that a word refers to 

• Proper nouns have a single referent (President Joe Biden refers to Joe Biden, the president) 

• Common nouns reder to many things (mug, table, chair) 

• Things that do not exist have no referent (unicorn, 4-sided triangles) 

• Different words/linguistic expressions can refer to the same thing. They can have the same reference 

  • Lin-Manuel Miranda/wtiter of the hit Broadway musical Hamilton… etc. 

Sense

• Sense/connotation - mental representation of a linguistic expression’s meaning

  • Definition/defining characteristics

  • Associations

  • Experiences

Example: Book

• Reference: set of all objects containing bound pages

• Sense: object collected in libraries, education, for knowledge and/or entertainment…etc. 

Example: stingy

• Reference: someone who doesn’t spend much money 

• Sense: not generous in a negative way 

Example: Frugal

• Reference: Someone who doesn’t spend much money

• Sense: someone who saves money in a positive way

How do sense and reference relate to each other? 

• An expression can have a sense without any referent 

• The same referent can be associated with different senses

• Reference is a property of some parts of language that relates the linguistic expression to things in the world 

Section 4: Thematic roles and primitive conceptual categories

Thematics roles

• Semantics of verbs 

• Describes the role that entities play in an action or relation

Agent - the individual which performs an action (the do-er)

Patient - something which is acted upon as part of an action

Theme - something which moves, literally or metaphorically, as part of an action

Source - the location/individual from which movement occurs 

Goal - the location/individual to which movement occurs

Location - the location at which something happens

Experiencer - someone who experiences something

Instrument - something an agent uses to make something happen

Cause - something that causes something to happen

Stimulus - something that causes an experience 

• Even though word order is different in different languages, they have the same thematic roles, which is able to cut through other linguistic differences between languages 

How do we understand the behaviour of words

• Some work in semantics has tried to understand the behaviour of words in terms of a set of primitive conceptual categories 

  • The basic building blocks of meaning

  • The semantics of every word can be decomposed into these elements 

Talmy

• Primitive conceptual categories → manner and path 

Manner: describes the specific motion encoded by a verb

Path: describes a physical or abstract trajectory

• Verbs encode manner or path information 

Manner = hopped, skipped, jumped, ran, limped

Path = enter, exit, ascend, descend

Typology

Languages tend to lexicalise (encode the meanings of) either path or manner in verbs (not binary but more focused on tendency/frequency)

Manner of motion - English, Greek, Slavic languages 

Path - romance languages, Turkish 

• When a language lexicalises one kind of information, that information is carried out by the verb 

  • The other kind of information (manner/path) is expressed by other means (usually via prepositional phrases) 

Example: In English, we use manner as verbs, and path is something that is added later 

• To walk (manner) down (path) 

However, in French, they use path as verbs, and manner is added in later 

• Je parti (path) au pied (manner) 

Talmy, Jackendoff

Primitive conceptual categories 

Manner: describes a physical or abstract trajectory

• From (start) 

• To (end) 

• Via (path passes near place) 

Pass

• Only means something that has happened in the past, behind… 

• You can say “the train passed through the tunnel” but you can’t say “the train passed to the station” 

• The verb ‘passed’ requires a via because it needs to be near something, but that is not the final destination

Climb

• Climb is a verb that means it’s towards something, or on top of something 

• ‘I climbed to the station’ works but ‘I passed to the station’ does not 

Generality of conceptual primitives

• Rent, lend, borrow 

  • A transfer of possession, specifying a theme (something which moves) and a path of possession

Maria lent the book to Julia 

• From (Maria) 

• Path (go, lend) 

• To (Julia)

Compositional Semantics

Compositionality

Principle of compositionality - the meaning of a linguistic expression is a systematic function of the meanings of its parts (lexical semantics) and their syntactic organisation

• An expression’s meaning is built from the meaning of its parts

• Powerful! This is how we understand utterances we have never heard before

Compositional expressions

• Words have meaning, and when they’re put together in a sentence you can take the individual meanings of the words and then string them together 

  • All ice cream is delicious = every instance (all) of a frozen sweetened cream product (ice cream) is gustatorily pleasing (delicious) 

• However, syntax matters! “Dog bites cow” is different from “cow bites dog” 

  • Who is doing the biting is determined by who is the agent which in English is determined by word order

Non-compositionality

• The meaning of some expressions is not compositional, meaning that the overall meaning is not a systemic meaning of its parts 

  • It’s raining cats and dogs = not raining literal animals, but raining very hard 

  • Break a leg = not actually breaking legs, but means good luck 

• In the case of non-compositional semantics, speakers have to memorise the holistic meaning rather than deriving it from its parts 

How does compositionality work?

Sets

Mathematical concept: a set - an orderless collection of entities

• Sets are well-defined, meaning we always know whether an entity is contained or not contained in a given set (no ambiguity) 

• They are also theoretical constructs, meaning we aren’t interested in whether we could actually/practically find all of the entities that should belong in a set (ex. Finding all of the mice in the world would not be practical)

Lexical semantics: reference of individual words

• The reference of individual words can be thought of as sets of entities 

Proper nouns - the reference of proper nouns can be thought of as a set containing a single individual (ex: Barack Obama = refers to Barack Obama, and not anybody else) 

Common nouns - nouns refer to the set of all entities that have a particular property (ex. Ice - the set of all things that are water and solid)

Verbs - verbs refer to the set of all events in the world that have a particular property (ex. To freeze: the set of occurrences where a liquid becomes a solid, to slaughter: the set of occurrences where an animal is killed for food) 

How does compositionality work?

How can we use sets to help us understand the meaning of compositional expressions? 

Sets: NP VP, Adj NP…

NP VP Assertions

Ordinary declarative sentences assert a proposition

Propositions have 

• Truth values - is the fact that is being asserted true? 

• Truth conditions - the conditions that must hold in the world for the proposition to be true

Truth values

• Propositions have truth values, meaning that propositions can be true or false 

• Nouns, verbs, etc… do not have truth values 

Example: In 2020 Tufts University had exactly 500 enrolled student (not true) 

• However, the words ‘enrolled students’ have no truth value

Truth conditions

The conditions that must hold in the world for a particular proposition to be true 

Proposition: sometime in the future, another world war will occue 

Truth condition: At least one time that is later than the time of the utterance, all/most of the countries in the world will be fighting with each other 

Proposition 2: There will never be another world war 

Truth condition 2: At none of the times occurring after the time of the utterance will all/most of the countries in the world be fighting with each other

• Ordinary declarative sentences assert a proposition 

  • Example: Pedro Pascal acts = asserts that Pedro Pascal is in the set of entities that act

Propositions

Declarative sentences don’t simply refer, they assert a fact about the world 

• This is the best group of Intro to Linguistics students ever! 

• Noun phrase - refer 

  • The sentence is asserting that the referents of the noun phrase have a particular property (are in a set)

ADJ NP - Specifying

How do adjectives modify NPs? 

• At a basic level, adjectives specify some subset of the entities referred to by the NP (brown cows are cows that have brown hair) 

• How they do this depends on the type of adjective

ADJ NP - Intersection

Some adjectives specify a subset of the NP by intersecting two sets 

• The ADJ specifies a set of entities

• The NP specifies a set of entities 

ADJ NP = the set of entities that are in both the ADJ set and the NP set

Example: Married felon = the set of all individuals who both have been convicted of a felony crime and are married 

• ADJ NP are intersective adjectives that independently specify a set of referents 

• Examples: Crime-fighting, Canadian, Drink, retired

ADJ NP - Relative and subsective intersection

• A big shrimp is considered small relative to a small elephant 

ADJ NP - Non-intersection

• Expression refers to individuals who may or may not be in the set referred to by the NP

  • Possible, likely, alleged, supposed

ADJ NP - Anti-intersection

• Expression refers to individuals that are not described by the NP

  • Examples: fake, pretend, artificial, imaginary

• ADJ entails entity is not in the set of items picked out by NP

  • A fake gun is something that is not in the set of things referred to by a gun (fake gun ≠ gun) 

• A sword isn’t a possible fake gun 

• Anti-intersection just means that it’s not something, not the set of all things not in a certain set

Pragmatics 1

Semantics

“Solina was hired by Moderna” 

Truth value = either true or false 

Truth conditions = Moderna extended a job offer to Solina, and Solina accepted 

• Semantics studies the literal meaning of language without context 

What is pragmatics?

• The study of how we use language in context, and how context contributes to meaning 

Key concepts:

• Meaning is context-dependent

• Conversations are structured and follow rules

• We makes inferences from language 

• People use language to do things 

Language in context

1. Context dependence

What a sentence means and what a sentence does depends on the context

Example: Context dependence 

A: How are you going to pay for the ticket? 

B: I have a credit card (Implicature: I will be paying for the ticket with my credit card)

A: I can lend you money for the ticket

B: I have a credit card (Implicature: No, I don’t need money because I have a credit card)

A: I don’t have any money to pay for the ticket

B: I have a credit card (Implicature: I will pay for the ticket using a credit card)

Continuum of Context Dependence

Context independent - The first president of the United States (is always going to be the first president, nobody else)

Context-dependent - deictic terms (pronouns, that, this… etc.) 

Deitic terms

Linguistic expressions that by themselves don’t mean anything specific. Their meaning is entirely dependent on the context in which they appear 

• Pronouns

• Temporal terms 

• Directional terms 

Example: “she is there now” could mean 

• A friend of yours is in class 

• Britney Spears is in the store 

• The tooth fairy is in your childhood bedroom

Different types of context

Linguistic - what has been said earlier in the conversation

Extra-linguistic: Context outside of language 

• Situational - the local environment of the utterance 

• Social - the relationships and norms of the social situation

Given vs. New Information

In utterances, informations can be given or new 

Given - information that is assumed by the speaker to be known or inferable by the addressee at the time of the speaker’s utterance, because it is 

• Common knowledge

• Part of the extralinguistic context

• Previously established in the discourse

New - information that is assumed by the speaker to be unknown or not likely to be inferred by the addressee

Example: Pumpkins for 1.99, with free pumpkin seeds inside (this is redundant because pumpkins come with pumpkin seeds)

Expression of given/new information

• New information tends to be expressed with full phrases 

• Given information tends to be referred to with single words and deictic terms 

Example: My roommates Patricia and Gavin spent the whole summer at their cabin in Maine 

• “While they were there, they swam in the ocean every day” 

• “They said it was really nice”

Context-dependence summary

• Linguistic expression vary in the amount of their meaning that is determined by context 

• Deictic terms have almost all meaning determined by contact 

• Context can be discourse, environment, knowledge, etc.

• How we formulate an utterance in conversation depends on whether the information is given or new

Section 2: Language in Conversation

Rules for Conversation

People infer meaning from utterances that is not contained in the literal meaning

• How do we do this? 

  • Language is governed by social rules that allow our communication to work

Grice’s cooperative principle

Paul Grice: a mental framework exists that allows listeners to make inferences 

Cooperative principle

• People in a conversation assume that what people say is intended to contribute to the purposes of the conversation. We assume people are being cooperative

Gricean Maxims

Assumptions that we hold when we are engaging in conversation

• Rules of thumb for how interlocutors convey information in conversation

• States prescriptively in form but are meant as descriptive statements 

• Not necessarily true across all cultures/contexts

Maxims of quality

• Expectations of honesty in conversations 

  • Don’t say what you believe to be false

  • Don’t say that for which you lack adequate evidence

Maxim of relevance

• Contribute to the topic, not get off-topic 

Maxim of quantity

• Provide the right amount of information

  • Make your contribution as informative as is required 

  • Do not make your contribution more informative than is required 

Maxim of Manner

• Make you contributions in a straightforward, common way 

  • Avoid obscurity of expression

  • Avoid ambiguity

  • Be brief

  • Be orderly

Violating maxims

Violating a maxim is when someone does not conform to the requirement of a maxim

• Can be unintentional or intentional

Example 1: 

A:What would you like for lunch? 

B: When’s screen time? 

• Violates maxim of relevance because the response is not directly relevant to the question

Example 2:

A teenager who has been hanging out with their friends all day returns to the house 

A: Where have you been all day? 

T: At the library studying! 

• Violates maxim of quality because it’s not truthful

Flouting maxims

• When a speaker says something that in its most literal form appears to violate a maxim, but the listener is expected to understand the intended communication (due to shared knowledge of the maxims) 

A: I’m pretty sure the tiktok I’m about to post will go viral

B: Right, and I’m going to win the Nobel Prize 

• Flouting a maxim is still cooperating conversationally (they are flouting the maxim to be sarcastic or funny) 

Felicity

• Situational appropriateness 

Felicitous utterances are pragmatically well-formed: appropriate for the context and cooperative 

• Infelicitous utterances are inappropriate given the context or uncooperative (violate a maxim) 

• This is different from grammatical/ungrammatical

Examples of felicity

A: What kind of pet do you have? 

B: I have a cat named Blackberry 

• This conversation is felicitous 

A: What kind of pet do you have? 

B: I have an animal

• This conversation is not felicitous and violates the maxim of quantity

A: What kind of pet do you have? 

B: I love drinking coffee in the morning! 

• This conversation is not felicitous and violates the maxim of relevance

Conversational pragmatics is hard

Conforming to these maxims takes a lot of cognitive processing 

• Need to know the maxims

• Know conventions for different situations

• Model interlocutor’s mental state and world knowledge

• Use this info to make predictions

• Update this information on every turn of the conversation

• Have to do it fast!

Pragmatics 2 and discourse

Making inferences

• What conclusions do we make from an interlocutor’s utterance? 

  • Semantics: the content of a proposition, and entailments based on propositions 

  • Pragmatics: conversational implicatures

Entailments: facts that must also be true if a given proposition is true 

• If A then B (B is entailed by A)

• Only statements that are logically necessary

Proposition and entailments

Monique has a red car

Proposition: Monique owns a red car (core meaning conveyed by a sentence)

• The set of things that Monique owns includes a red car 

Entailments:

• Monique owns a car

• Monique owns things 

However, “Monique owns a sports car” is not entailed because it is not logically necessary 

Maria stole the diamond

Entailment: Maria is a thief 

Maria allegedly stole the diamond

• Does not entail that Maria is a thief 

What do we learn from a statement?

Semantics

• Proposition

• Entailments of the proposition

Pragmatics

• Implicature 

Inference

Inferences are conclusions that people draw from the state of their environment 

• May or may not be true

Conversational implicature

Implicature - inferences made based on language

• Specifically: Information inferred from an utterance that is not the literal proposition or an entailment of that proposition

  • The non-literal meaning of an utterance

  • Implying information through language 

• There are many different kinds of implicature

  • Sarcastic parody

  • Conventional phrases 

  • Using the maxims of conversation

Using the maxims to make an inference

Method 1: 

• Speaker conforms to the maxims, listener makes inferences from the maxims

Implicature based on the maxim of quality

A: Burt has a motorcycle

Implicature: A believes Burt has a motorcycle and has evidence for it

• Maxim of quality = only say that for which you have adequate evidence for 

Burt has a motorcycle but I’m not sure ≠ Burt has a motorcycle 

A1: Rebecca took the medication and had an allergic reaction 

Implicature: Rebecca had an allergic reaction to the medication

A2: Rebecca had an allergic reaction and took the medication

Implicature: Rebecca took the medication to treat the allergy

Scalar implicatures

• When dealing with quantities or quantified qualities, stating X implies nothing stronger (along the relevant dimension) is true 

Using the maxims to make inference

Method 1: Speaker conforms to the maxims, listener makes inferences from the maxims 

Method 2: Speaker flouts the maxims, listener assumes they are being cooperative, listener reinterprets the utterance

Revisiting flouting maxims 

Violating: one does not conform to the requirement of a maxim

Flouting: violating a maxim in such a way that it is obvious and intended to be discovered

Implicature based on the maxim of relevance

A: I’d really love a cup of coffee

B: There’s a place around the corner called Joe’s

• B blatantly violates (flouts) the maxim of relevance 

• A assumes B is being cooperative 

• A reinterprets B’s utterance to find the relevance 

Implicature: Joe’s sells coffee

Humour

Jokes often rely on violating/flouting maxims to create humour

• They work by violating pragmatic expectations

Section 2: Speech Acts

What is language for? 

• We use language to do a huge range of activities 

• With language, we perform actions

  • Request information

  • Give orders

  • Apologise

  • Give compliments

Doing things with language

1. John jones was at the office yesterday until 6 pm (assertion)

2. Who ate all the cookies (question)

3. Sit down and be quiet (command)

4. Please let me know if you’ll be attending (request)

5. If you do that again, I’ll report you (threat) 

6. Watch out – there’s a huge pothole there (warning) 

7. Give bucks says that the Buckeyes will beat the Wolverines this year (bet)

8. I really like your t-shirt (compliment) 

Common speech acts

Assertion - conveys information

Question - elicits information

Request - elicits action or information

Order - demands action

Promise - commits the speaker to an action

Threat - Commits the speaker to an action that the hearer doesn’t want

Felicity conditions for speech acts

In order to be felicitous, speech acts must be uttered in a certain kind of context. There is a set of conditions that must be true for a speech act to be felicitous 

For requests: 

• Speaker must believe that the action has not yet been done

• Speaker wants the action to be done

• Speaker believes the hearer can perform the action

• Speaker believes the hearer may be willing to perform the action

For question: 

• The speaker does not know some piece of information

• The speaker wants to know that information

• The speaker believes that the hearer may be able to supply the information

Performative speech acts

Utilise specialised verbs to accomplish the speech act denoted by the verb

• I advise you to buy that house

• I order you to fire that gun 

• I assert that the following testimony is true

• I warn you to go no further

• I promise I’ll stop bothering him

Ceremonies and performative actions

Certain ceremonies require the use of performative actions

• I hereby pronounce you husband and wife

These have felicity conditions usually associated with the authority of the speaker, sincerity of participants

Summary

Pragmatics is the study of how we use language in context, and how context contributes to meaning

Key concepts: 

• Meaning is context-dependent

• Conversations are structured and follow rules

• We make inferences from language

• People use language to do things

Conversational pragmatics is hard!

Conforming to these maxims takes a lot of cognitive processing: 

• Know the maxims

• Know conventions for different situations

• Model interlocutor’s mental state and world knowledge

• Use this information to make predictions

• Update this information on every turn of the conversation

• Have to do it fast!

Language acquisition

Our linguistic behaviour is quite complex, but babies aren’t born talking. Studying language acquisition can reveal

• Something about the nature of language itself 

• Something about the part of the mind responsible for language

Summary

• Babies are linguistic geniuses. Thousands of linguists have been trying to figure out linguistics, but babies acquire it unconsciously by the age of 10 or so, and unaided as well 

Section 1: Milestones in language acquisition

Progression

Language development generally proceeds according to a set of milestones 

• Acquisition proceeds at a regular pace

• In a particular sequence 

• No matter which language the child is learning

• Children comprehend more than they can produce at first 

0-1 month - cooing (they make noise that has phonology, but not words)

6 months - babbling 

• Infants begin to produce consonant-vowel sequences with segments from both ambient and non ambient languages 

• Some sounds are more likely to be produced than others (plosives are more likely, fricatives are less because they are more complicated) 

8-11 months - first signs of word comprehension (recognise words like ‘eat’)

11-13 months - first word production

12-18 months - one word speech, and accumulate 50-200 words 

• One word stage: single word utterances used to convey a variety of message types 

• Daddy = father enters room

• Juice = asking for juice 

• Blanket = pulling blanket away from sibling

• They don’t have the words to convey what they want with the words, but they can make simple words 

• Also have phonological simplifications 

  • Syllable deletion = usually only stressed syllables are pronounces

  • Fricatives and affricates = are difficult to pronounce, so they’re changed to stops

  • Loss of codas (end of word)

  • Cluster simplification + coda loss (loss of beginning and end of word)

  • Velar → alveolar fronting (goose = doos)

  • Consonant harmony = rhyme things (butter = bubba)

18-22 months - vocabulary spurt and two-word speech 

• Baby chair = the baby is sitting in the chair (have semantic relations. Theme → location)

• Doggie bark = the dog is barking (agent - action)

• Ken water = Ken is drinking water/wants water (agent - theme)

• Hit doggie = I hit the dog (action - theme) 

2-4 years - more complex syntactic structures 

• Telegraphic stage (22-24 months) where there is generally correct syntactic order, and speech is primarily made up of content words and few function words 

• Chair all broken = The chair is all broken (function words are ‘the’ and ‘is’)

5-6 years - most basic syntactic structures, know about 12-14k words, and fluent production of speech 

Development

1. Comprehension precedes production

2. All linguistic levels develop simultaneously

3. Competence and performance develop simultaneously 

• Immature language can be due to 

  • Immature competence (their brains don’t remember)

  • Immature performance (they physically can’t figure out how to pronounce the word) 

  • Both 

Example: Why are the inflections and function words omitted during the telegraphic stage? 

Competence: Child has not yet acquired a full morphosyntactic grammar 

Performance: Immature phonological abilities may limit input and output abilities

• Comprehension

  • Inflections are often single segments and function words are phonetically reduced in speech

  • May be harder to learn inflections and function words from the signal 

• Production

  • English inflections are often single segments and form clusters 

  • Children may not be able to articulate inflections if they cannot produce clusters 

  • Walked sounds a lot like walk (in regular speech, we don’t tend to exaggerate the -ed) 

Many areas of development

• Morphological and syntactic grammars are still developing (competence) 

• Phonological limitations (performance)

Summary of Section 1

• Language acquisition typically proceeds through a series of stages 

• Comprehension leads production

• Errors systematically reflect stages of development (e.g. phonological simplifications) 

• All aspects of language develop alongside each other 

• Competence and Performance develop alongside each other 

Section 2: How children learn

How do children learn language? 

Common belief: parents and schools teach children language 

• However, babies receive positive evidence (sentences consistent with the grammar to be learned when spoken to) 

• They do not receive negative evidence (that’s wrong) or instruction (this is how you form a question)

Brown and Hanlon (1970)

1. Transcripts of parent-child interactions

2. Categorised child’s productions as grammatical or ungrammatical

3. Checked whether the parent expressed approval or disapproval

Found out that the parental approval/disapproval did not differ in response to grammatical and ungrammatical utterances 

• When teachers/parents teach children, they can pronounce the correct word, but they don’t integrate it into their sentences 

Section 3: Children acquire grammars

Question

Children are learning all sorts of things, including memorising words. We know that children are also acquiring grammars 

• We know this through productivity (remember, language is productive, yay!) 

Novel Utterances

• I hate you, and I won’t unhate you 

• My teacher holded the baby rabbits

Nobody has told them these words, but they have learned English grammar and are applying it, even if it’s not right in certain environments. This is called over-regularisation

Over-regularisation

• The application of grammatical rules to an irregular item

• This provides strong evidence for having learned a rule 

Section 2 and 3 Summary

• Children receive positive evidence and little negative evidence/instruction (instruction is ignored) 

• Children induce productive patterns (grammars) 

• Errors frequently reflect over-application of rules

Section 4: Contributions of experience and innateness to language acquisition

Experience Matters

• Children learn the language they are exposed to 

• Languages have changed dramatically over just 1000 years 

Therefore, experience plays a crucial role in language acquisition, and language learning is necessarily experience-dependent (we aren’t born with a language programmed in us)

Contribution of innateness

• Cats don’t learn language, and experience is necessary to learn a language. But is it sufficient?

• Humans are the only species that have language

• Have we evolved to produce and understand language? 

Nativism - there are neural structures that are specifically involved in language acquisition

Today, some parts of an argument for a contribution of innateness

1. Biological aspects of language acquisition 

2. Poverty of the stimulus

Biological aspects of language acquisition

Innately determined behaviours

Research done by Lenneberg in 1967

• Present in all typical individuals of a species

• Not the result of a conscious decision

• Direct teaching and intensive practise have relatively little effect

• Regular sequence of milestones as behaviour develops

• Likely to be a critical period during which the behaviour must be acquired

Critical period

A period of time in development when input from the environment is necessary for typical functioning to develop

• Happens in vision = need to see vertical lines for binocular depth perception

• Also happens in language. There are a few cases of linguistic input absent during early life (greater than 12 years)

  • These people never developed full language abilities (particularly syntactic) despite adequate later exposure

  • This suggests that there is a critical period for first language acquisition, perhaps ending around puberty

Genie

• Isolated since 20 months, and rarely, if ever, spoken to 

• Found at 13 years old

  • Although after she was found, she was taught language, her vocabulary grew but her syntax did not 

• Applesauce buy store = buy applesauce at the store 

  • Missing function words

Chelsea

• Born deaf, never diagnosed as deaf and wasn’t able to learn sign language 

• She was first fitted with hearing aids and exposed to speech at 32 

Following exposure, her language performance was still limited

• After 12 years, her vocabulary was around 9-10 grade 

• Lack of syntax

In both cases

• Abnormal linguistic input early in life

• Despite much subsequent exposure, syntactic ability never approach that of a typical adult

• Provides evidence for a critical period for language 

• Suggests language learning has a biological (innate) component

Evidence for a contribution of innateness

1. Biological aspects of language acquisition

   1. Language bears the hallmark of a biologically driven behaviour

   2. Language development appears to include a critical period

2. Poverty of the stimulus

Poverty of the stimulus

• In cases where the linguistic input is impoverished in some way, children learn linguistic grammars that are more sophisticated, more complex, more language-like than what they are exposed to 

Nicaraguan Sign Language

• First cohort was children 10 or older, and second cohort was children who joined later at 4 years of age

• First cohort developed a signing system that was like a pidgin, and the second cohort had more sophisticated constructions 

  • Their signing increased in fluidity, was more compact and stylised

  • Grammaticalisaiton - had a point in space for an agent, and a point in space for theme

It’s not the oldest children who are the most advanced, it’s the youngest. They seem to take in the input that is not sophisticated and to regularise it, make it more complex

Poverty of the stimulus summary

• Young children who receive impoverished input regularise and improve upon their input

• If language learning were 100% experienced-based, these children should match their input

• The fact that they make it more complex suggest they bring something innate to the table when learning languages (they’re deaf)

Experience vs. innateness summary

• Language acquisition is clearly experience-dependent 

• However, evidence from critical periods and poverty of the stimulus suggest that language acquisition is aided by neural structures that are specifically related to language

Language change

Terminology

Synchronic - pertaining to language at a particular point in time 

• Synchronic linguistics is interested in characterising the mental grammars of speakers based on the structure of the language at the time they learn it 

Diachronic - pertaining to language across time 

• Diachronic linguistics is interested in characterising how and why languages change over time 

Historical linguistics - the branch of linguistics that studies language change

Language Change

Language change occurs at all levels of linguistic structure 

• Phonology

• Morphology

• Lexical (vocabulary)

• Syntax

• Semantics

• Pragmatics

Why do languages change?

• Creativity

• Social factors

• Language contract

• Geographic division

• Factors that lead to variation in what is learned

Section 2: Word-level change

1. Adding words (non-morphological means) 

2. Semantic change

3. Word choice

4. Word-specific phonological change

Non-morphological ways of adding words (language internal)

Acronym - combining initial letters or sounds of a phrase into a pronounceable word

• Care package = cooperative for assistance and relief everywhere

Blend - combining the phonemes of two or more words 

• Portmanteau, snowmageddon (snow + armageddon) 

Clipping - shortening of a word 

• Application = app

Coining - creating brand new words by combining phonemes 

• Quark, Bumble… etc.

Adding words (language external)

Borrowing - taking a word from another language and changing it so that it fits into the new language’s syntax 

• Cookie came from Dutch, meaning little cake

Semantic changes

• Words are continuously changing their meaning

  • What they refer to (reference) 

  • Their connotations (sense) 

• Words also change in their social status

  • Meat = solid food, now means animal flesh 

  • Naughty = meant poor or needy, now means poorly behaved

  • Bully = meant sweetheart, darling, now means person who seeks to harm

Semantic extensions and reductions

Pertains to changes in the number of things a word refers to and how specific/general it is 

Broadening/extension 

• The meaning of a word broadens (cupboard = drawer for cups/dishes. Now means regular cabinet for things in kitchen)

• The number of things it refers to increases 

• Becomes less specific, more general 

Reduction

• The meaning of a word narrows 

• The number of things it refers to decreases, becomes more specific, less general (meat used to refer to all solid food, and now only refers to animal flesh)

Semantic elevations and degradations

Pertains to changes in the sense/connotations of a word 

Elevation 

• Word becomes more positive (pretty used to mean cunning, and now it means attractive)

Degradation

• Word becomes more negative (accident used to mean a chance event, and now it means an unfortunate occurence) 

Word choice

What word we use to express a concept changes over time

• Living room = parlor, drawing room, sitting room 

• To date someone = to court someone 

What influences word choice?

• Why not use multiple alternative words? 

Vocabulary convergence - natural tendency to express a concept with a single word 

• Social cachet

• Cross-generational opacity

• Formal considerations 

  • Brevity

  • Phonotactic frequency

  • Lexical support

A noisy serenade used to teased married couples had many different words for it depending on region, but it settled on Shivaree 

• Because the final stressed [i] sound were also found in words with similar festive connotation (whoopee, jubilee)

• So shivaree received lexical support while callathump did not

Word-specific phonological change

Sometimes the phonological form of specific words will change their phonological form 

• Cupboard used to sound like cup board, and now sounds like cuboard (because the pb is hard to pronounce together) 

• Police used to have a different stress (POlice and poLICE) because it was borrowed from French

In many cases, this is the result of phonological regularisation 

• An alteration to a word’s form to improve its phonotactic/phonological well-formedness

• Similar to loan word adaptation

Section 3: Morphological change

Word specific vs. language-wide change

• Changes to specific words involve changing specific lexical entries

• Changes that are language-wide involve changes to grammars

When enough changes to specific words accumulate and align, this can lead language users to infer a new grammatical rule

Paradigm levelling

• A paradigm is the set of inflected forms of a word 

• Paradigm levelling is the complete or partial elimination of variation (allomorphy) within a paradigm 

• This is an instance of morphological regularisation

• A pressure/preference for uniformity causes idiosyncracies to become erased in favour of a dominant pattern

Language-internally driven processes

Sometimes, affixes originate in old words. Modern English -ly descends from Old English lice (meaning body)

• Compound mann-lice = having the body or appearance of a man

• - lice was used in so many compounds it 

  • Lost its stress and final consonant

  • Had its meaning generalised to “pertaining to”

Language-externally driven processes

• Old English did not contain the suffix -ment 

• Middle English borrowed a large number of verbs and nouns derived from those verbs ending in -ment from Old French and Anglo-French

  • Achieve/achievement, commence/commencement, judge/judgement

  • These were just stored as whole words 

• Speakers eventually inferred a productive rule of -ment suffixation

Section 4: Sound change

Phonetic change

• A change to the physical properties of an allophone (how it is pronounces) 

• Creation/change to a phoneme’s allophones 

By itself, phonetic changes do not necessarily result in changes to the system of contrast 

• Ow fronting is an example this. Ow = ew

Phonetic change changes how a phoneme/allophone is phonetically realised but by itself does not result in phonological reorganisation

• It does not change the system of contrast 

Phonological change

Phonetic change - change to how a phoneme/allophone is phonetically realised 

Phonological change - a change to the system of contrast in a language, a change to the mapping between allophones and phonemes 

English low-back vowel merger

A phonetic process has caused [a] to move towards [c] in many varieties of North American and Singaporean English

• Dawn = Don

• Caught = Cot 

There is not discernible difference between the two because the two vowels have merged together. They are not minimal pairs anymore

• This leads to the loss of contrast, and speakers of these varieties have 1 phoneme instead of 2

English /f/ Phonological change

• English [f] and [v] were two allophones, and [v] was only pronounced between voiced phonemes 

• However, words were borrowed from French that had [v] in non-triggering environments, which created a contrast with the voiceless [v]

• Because of this, english developed a /v/ and now English speakers have inferred that [f] and [v] are allophones of different phonemes 

  • Fine ≠ vine

Section 5: Sources of language change

What causes word-level change?

• Miscellaneous (semantic, social…)

• Regularisation 

  • pressure/support from a lexicon (a.k.a. Existing words, example: shivaree)

  • Pressure from a grammatical rule (cupboard)

What causes grammatical change?

• Speakers learn morphological and phonological rules by extracting patterns present in the lexicon (kick → kicked, laugh → laughed… and so on)

• Grammatic change is caused by a change to the synchronic grammar, which changes the linguistic data (or vice versa) 

• If enough words in the lexicon do contain a new pattern, speakers will learn/infer it as a productive rule

  • Therefore, if words in the lexicon do not contain a pattern, speakers will not learn/infer it as a productive rule 

  • The lack of minimal pairs in Singaporean English doesn’t do much, and the outcome is only one lost phoneme (not a new -ly, or a distinction between [f] and [v] like in other cases) 

Summary

• Language learning is not transplanting words and rules from one brain to another 

• If it were, language would not change (we learn language from others and because we do that, it changes slightly over time. Things are not perfectly replicated) 

• Our grammars will influence what we say, and what we say will influence what grammars we learned 

  • Some patterns will be diluted, and some patterns will be added 

  • Pressure like uniformity will influence this 

Sociolinguistic 1 - Variation

Overview

So far we have generally been ignoring variation

• We acknowledged that languages (grammars) vary geographically but it hasn’t been our area of interest 

• Otherwise we have treated speakers as homogenous within a specified region

  • Also have treated speakers as homogenous unto themselves (someone always speaks the same way)

These have been some of the assumptions of traditional generative linguistics, but are not true

Setting the stage

Sociolinguistics is the study of linguistic variation and how language use variables by and signals social identity

• How does language use vary geographically? 

• How does language use vary based on social identity? 

• How does language use vary by context?

Sociolinguistic factors

• Geography

• Socioeconomic status

• Age

• Sex

• Gender identity

• Sexual orientation

• Social contexts and speech tasks

• Linguistic environments

• Cognitive factors

Terminology

Rhoticity - whether a variety contains an /r/-like consonant in syllable coda position

• Rhotic = a variety that contains an /r/ 

• Non-rhotic = a variety that does not contain an /r/

In non-rhotic dialects, words that historically have an /r/

• Have no consonant there (car = ca) 

• Or may have a neutral vowel in its place (here= Hee_uh)

Section 1: Causes of Geographic Variation

• Also known as dialectology, the oldest and most studied form of variation 

• How do specific varieties differ from each other

• Why does geography affect language use?

Why does geography affect language use?

1. Language contact 

   1. Settlement patterns vary by location 

   2. Layers upon layers throughout time

Example: US settlement from Europe 

Early settlers came from different locations in Europe and brought slaves from Africa 

• Geographical ‘seeding’ of grammatical features from different English varieties

• Geographically-specific contact with different languages 

Westward expansion blended these features, so that is why there are more distinct accents in the East compared to the West 

Early US Settlement

Jamestown, Virginia (1607)

• Many settlers from Southeast England (London) 

• /r/-lessness was becoming common in Southwest England and was becoming the prestige variety

Boston

• Many /r/-less settlers from Southeast England 

Philadelphia/Midland region (1680s)

• Quakers from Northern England (/r/-full)

• Settlers from Wales and Germany

• Scots-Irish

Charleston (1670)

• Heterogenous English Varieties: English, Irish, Welsh

• Huguenots (France), Dutch

• Slaves from West Coast of Africa, Caribbean (Barbados) 

New Orleans (1717)

• French (and some German)

• Slaves from West Africa, Caribbean

• Acadians (Cajuns) from Canada

Westward Expansion

People from all regions moved west, blurring dialect boundaries 

Northwest 

• British followed by New Englanders 

• Later: Ohio valley, Tennessee

Southwest

• First: Spanish speakers 

• Settlers from many parts of the US

• Settlers from Asia

Why does geography affect language use?

1. Language contact 

   1. Settlement patterns vary by location

   2. Layers upon layers throughout time 

2. Language change 

   1. Natural barriers allow and limit contact 

Language change: Isolation and divergence

When a language community does not have contact with other speakers, change can occur 

Divergence

• Older structures (eg words, phonological patterns) may be retained in one community and not the other

• Changes/innovations may develop in one community and not the other

Isolation and Divergence

Physical Isolation

• Settlers arrived around 1629 from Cornwall and Devon 

• Relatively isolated up through the beginning of the 21 century

  • Only accessible by boat 

• Some changes on the mainland haven’t propagated to the island 

• Changes on the island haven’t propagated to the mainland 

  • Phonological patterns

  • Vocabulary about catching crabs

Section 2: Examples of variation in US dialects

Variation

Varieties differ from each other at all linguistic levels

• phonetics/phonology

• Vocabulary

• Morphology

• Syntax

• Semantics

• Pragmatics

Examples: lexical variation, semantics, syntax, phonology

Lexical variation

Varieties may use different words to mean the same thing 

• Sneakers = tennis shoes = gym shoes =running shoes

• Green beans = string beans = snap beans 

Variation in meaning

Broadening - a word is used to refer to more things than before 

• Barn - UK: a building meant for storing grain, US: a building for storing all farm-related items 

Narrowing - a word is used to refer to fewer things than before 

• Corn - UK: any type of grain, US: maize 

Non-standard syntactic constructions

• Needs VERB = the car needs washed 

• Wants PREP = the dog wants in

• ‘Positive anymore’ = He’s handsome anymore

• Done NP - Don’t talk to me until I’m done my coffee

Done NP

• I’m done my homework

• I’m finished my homework 

Said in Philadelphia, Vermont, Canada

• Different from “I have done all my homework” and “I am done with my homework” 

• Similar to “ready for school” since both can be modified by all 

  • I’m all done my homework 

  • I’m all ready for school 

Variation in sound

1. Phonetic 

   1. Differences in pronunciation of individual phonemes 

   2. Chain-shifts (inventory-wide changes to pronunciation) 

2. Phonology

   1. Differences in Inventory

   2. Differences in rules 

Phonetic Differences

The same phoneme can be pronounced differently in different varieties

• Time (Southern = tahm)

• Down (Philadelphia = deuwn)

Chain Shift

Many vowels in a variety change their pronunciation

• Change in one vowel leads to change in the next

• This allows the vowels to remain contrastive 

Northern Cities Vowel Shift

1. Caught vowels lowers front to Lot 

2. Lot vowel fronts closer to Trap

3. Trap vowel moves to Dress/Face

4. Kit, Dress, and Strut vowels move backward 

As one phoneme’s pronunciation moves into the space of another phoneme’s pronunciation, the latter’s pronunciation shifts as well. 

• This maintains phonological contrast 

Variation in Sound

1. Phonetic 

   1. Differences in pronunciation of individual phonemes 

   2. Chain-shifts (inventory-wide changes to pronunciation) 

2. Phonology

   1. Differences in Inventory

   2. Differences in rules 

Differences in Inventory - Additional Phonemes

• Scottish English voiceless velar fricative [x] 

  • Loch 

Additional phonemes lead to more contrasts 

• US english has lock and log (ck and g) 

• Scottish English has Lock, Log, and Loch (ck, g, and x)

  • Throaty sounding 

Fewer phonemes

US low back merger 

• Caught = cot 

This leads to fewer contrasts

Differences in phonological rules

Pin-Pen Merger

• Low back merger occurs in every word 

• Pin-Pen Merger is conditioned by the environment 

• Rule: /e/ → [i]/_[+nasal]

Sociolinguistics 2

Section 1: Variation by social identity

Speech style - linguistic variation linked to social identity

• How we speak marks us with respect to our social identities

• We use language to construct our identities

• Speech varies by social status, gender identity, racial identity, religion, ethnic identity, sexual orientation… etc. 

NYC rhoticity

• Prior to WWII, non-rhoticity was prestigious

• After WWII rhoticity became the prestige norm

Socioeconomic status

• Labov (1962) 

• Listened to speech from 81 adults from the Lower East Side of NYC

  • They were asked to pronounce words with /r/ in coda position given a variety of tasks 

• He found that the rate of /r/-production correlated with socioeconomic status 

• Casual speech = upper middle were more rhotic 

• Rhoticity rose when people were asked to make minimal pairs, word lists 

Age and gender

• Found that younger women in Vancouver and Toronto fronted their ‘aw’ much more often than males in their age group, and much more often than older women

Investigating gay speech

Beliefs/stereotypes

• There is a gay speech style 

• Gay men lisp

• Gay men have effeminate voices

• Lesbians have ‘butch’ voices

Pirrehumbert et al (2004) Gay Speech

• Recorded women from self-identified gay and straight men and women 

  • Had them say sentences

Results

• Naïve raters rated the recordings on a scale of 1-7 (where 1 was totally straight, and 7 was gay/lesbian) 

• The results: 

  • Straight men: 3.2

  • Gay men: 4.6

  • Straight women: 3.2

  • Lesbian/Bi women: 4.3

• These results were statistically significant and shows that there is a speech difference between straight and gay people

Acoustic Analyses

• Lesbian/Bi woman pronounced their /a/ and /u/ farther back than straight women 

• In gay men, they had a larger vowel pronounciation space compared to straight men (their fronting was more front, their back vowels were more back… just bigger space overall) 

However

• Lesbian/Bi women did not mimic vowel space of men, and gay men did not differ in pitch or breathy voice

Conclusions

• There does appear to be a gay speech style

• Listeners are capable of discerning these cues

• Different for lesbians/bis and gay men

• Does not involve mimicry of the other sex

Social status

Norma Mendoza-Denton did study on California Latina Gangs and studied how elements of California gang’ members speech, bodily practices (tattoos) and symbolic exchanges signal their affiliations and identities 

• People who were in a lower level in the gang spoke differently compared to people higher up 

California rising

As part of the California Vowel Shift, [I] is probabilistically raised to [i]

• The rate of California raising correlated with status within the hang 

Section 2: Variation by social context

Time Scales

Geographic and identity-based patterns to language use are fairly stable over time

• However, language use can also vary on faster timescales 

  • Speech task

  • Social context

Effect of speech task

• Labov (1962)

  • Found that the production of /r/ (rhoticity) varied as a function of the speech taks

Style-switching

• A stylistic switch in language use to align with one’s context or with one’s perceived identity in a given situation

  • May be intentional, may be conscious

Register

• A collection of linguistic behaviours and features associated with a particular social situation

• What about the situation matters? 

  • Topic, purpose, modality (spoken/written), genre, social relations…

Guess the relationship/social situation

• Dear Professor Goldberg (formal email, student to teacher) 

• Hi Professor Goldberg (informal email/respectful speech, subordinate to higher ups_ 

• Hi Ariel (casual speech between equals) 

• Hi Ari (Casual speech between friends) 

• Hey Dummy (either really mean or they’re really close friends)

Joos (1961) 

Frozen - unchanging language (Pledge of Allegiance, poems, bible quotes) 

• Need to memorise language 

• Formulaic, possibly archaic structures 

• Interaction is either none or 100% prescribed

Formal - one-way participation among strangers (presentations) 

• No interruptions, complete sentences are required, word choice is important, and fewer contractions 

• Maximises social distance, background knowledge not assumed 

Consultative - two-way participation (Teacher/student, doctor/patient) 

• Background knowledge is not assumed, back-channelling is common (mm-hmm, uh-huh) 

• Establishes and maintains a neutral and task-oriented distance

Casual - conversation among in-group friends 

• Background information not provided, more general (non-technical_ vocabulary and interruptions are common

• Used to establish/maintain familiarity and shared experiences, attitudes, etc. 

• Back-channeling and interruptions are common

Intimate - family members, close friends 

• Word choice is less important than intonation, may include private vocabulary 

• Used to establish maximum intimacy among people who interact regularly

• Background knowledge is assumed 

• Back-chanelling and interruptions are common

Registers vary at all levels

Vocabulary/Morphology

• Addressing with names vs. titles 

• Use of technical vocabulary 

• Contractions vs full forms 

• Tabboo words 

• *got = as in I got this 

Syntax

• Active vs. Passive sentences 

• How 1st/2nd/3rd person is references 

• Degree of modification (ice cold shrimp nestled on fresh organic lettuce = shrimp on lettuce) 

Phonology

• “Standard” vs “Non-standard” pronunciations 

  • Library vs libry

Section 3: Theories of Variation

• How do we account for the ways in which we see language use vary? 

1. Variation as Grammar

2. Theories of Style Shifting 

   1. Attention to Speech

   2. Audience design (listener-oriented speech)

Variation as Grammar

Variation is systematic, so it makes sense to think of it as grammatical rules 

• Speakers of different varieties learn different grammatical rules

• Speakers learn different rules based on their social communities/affiliaitons

Example: Variable Non-Rhoticity

We can think of /r/-deletion as a grammatical process 

/r/ → ø/_(Coda)

• This means that /r/ is deleted when it is in coda position

Variable Rules

Variable rule - A grammatical rule may apply different proportions of the time (80% vs 20%)

• Rather than being deterministic (applying 100% of the time), rules will apply probabilistically 

• Not 0%-100%, but somewhere in between 

Variation as grammar

Theory: speakers of some varieties and socio-economic groups learn the /r/ deletion rule as applying more frequently than speakers of other varieties and groups 

• Speakers of Southern American English: Around 0% of the time 

• Speakers of NYC English: greater than 0% of the time 

• Speakers of higher socio-economic status do it less often than speakers of lower socio-economic status 

Within the same speaker, style switching (e.g. as a function of speech task) will involve temporarily changing the rate at which a variable rule applies 

Theories of style switching

Attention to speech

Style varies according to the amount of attention the speaker pays to their speech

• Ranged from unguarded, little attention (e.g. casual conversation) to full attention

Result

• /r/ varies with speech task 

• Tasks that involve more “attention” lead to pronunciation of /r/ more often (like minimal pairs)

Theories of Style Shifting

Audience Design

• Speakers adjust their speech according to the needs of and their relationship with their interlocutor 

Convergence

• A speaker will attempt to match the speech of their interlocutor. Signals solidarity, gains social approval

Divergence 

• A speaker will shift their speech away from that of their interlocutor. Signals social distance, perhaps disapproval

Convergence Example

Monopththongisation is common in Southern and Black varieties of English

• Hay et al (1999) explores how Oprah Winfrey’s speech vary according to the race of her guest through acoustic measurement of ‘I’

  • They found that Winfrey used more monophthongs in her introduction when she was introducing an African American guest than an American one

• This is an example of audience design

  • Speakers adjust their speech according to the needs of and their relationship with their interlocutor (Baese-Berk and Goldrick, 2009)

Voice Onset Time

• In English, the primary cue to stop voicing is Voice Onset Time

Voice Onset time - the length of the burst of air before voicing starts

Perception Difficulty

• Minimal pairs differing in voicing differ in their VOT

  • Example: Cod, god, 

• One way to be clearer about which word you’re saying is to hyperarticulate 

Hyperarticulation - larger, longer articulatory gestures 

• Increasing the VOT of Cod will increase the difference between Cod and God (ccccod, god)

Hyperarticulation of Voice Onset Time

• Baese-Berk and Goldrick (2009)

• 2 participants at a time, each had their own screen 

  • 3 words on both screens, speaker had to say which one to click

  • Had either minimal pairs absent or present 

Question: Do speakers accommodate the difficulty posed by the presence of the minimal pair? 

Results

• VOT was longer when the minimal pair was present onscreen

• VOT of cod is longer in the presence of god

Section 4: Social Aspects of Language Change

Sex and linguistic innovation

Women are often linguistic innovators, using a form that becomes dominant before the general population 

Example: Vancouver “aw” fronting

Linguistic Prejudice and Discrimination

Why are we here?

Language is 

• Extremely detailed, complex, and nuanced

• Intimately related to our identities

• Intertwined with every aspect of the human experience 

It’s natural that it would intersect with discrimination and prejudice (natural does not mean correct) 

To discuss how language intersects with discrimination and prejudice 

1. Language as a conduit for identity

2. Language as a basis for discrimination 

Many more other intersections such as educational policy, bias in natural language processing… etc.

Section 1: Discrimination where language is a conduit for identity

We can identify elements of a person’s identity through the way they use language. People are very good at identifying someone’s ethnicity based on how they sound (Purnell et al. 1999)

Fair Housing Act

In 1968, U.S. Federal legislation that protects individuals and families from discrimination in the sale, rental, financing, or advertising of housing 

• Failed to pass Congress for 2 years until the death of Martin Luther King

Housing Discrimination - Does language serve as a proxy for identity?

Anecdotally Black and Hispanic renters get fewer call-backs about housing opportunities, suggesting discrimination on the basis of voice alone

• However, the notion of discrimination based on voice has been questioned at the judicial level

Purnell et al conducted phone interviews with landlords, posing as a potential renter in 5 different Bay Area locales 

• He varied whether he used Standard American English, African American Vernacular English, and Chicano English

• Overall people who speak Standard American English get called back to areas that are white, Black neighbourhoods who have more black people will get a higher callback, but not in neighbourhoods with less black people, and same with Hispanic people

How sensitive are we to sociolinguistic cues to race?

Speakers of the 3 dialects recorded “Hello, I’m calling to see about the apartment you have advertised in the paper”

• Naïve raters heard each recording and had to decide ethnicity from the 3 options

• The results were statistically significant 

Landlords/sellers screen out unwanted applicants on the basis of their accent 

• This is discriminatory profiling: using language as a proxy for race, class, or national origin

• Under certain circumstances, identification is possible on the basis of just one word

Language is a conduit for identity

• Language → Infer identity → Discrimination 

Section 2: Linguistic Prejudice

Why are we here?

Language use at all levels is shaped by an individual’s social identities 

• Language users are extremely sensitive to socially-relevant linguistic properties

  • They also frequently have strong normative viewpoints about language use 

Linguistic Prejudice

• Ascribing negative (or positive) properties to people because of they way they speak/sign

• Linguistic prejudice can lead to negative outcomes in social, academic, workplace, and legal settings, potentially having severe consequences for marginalised and racialised people 

  • Usually implicit

  • Reinforced by neutral to well-meaning media

  • Usually viewed as a matter of “correctness”

• Linguistic prejudice is also one of the last “acceptable” prejudices

Women’s speech

• Tend to have vocal fry (Think Kim Kardashian) 

• Or uptalk, where their pitch rises at the end of a sentence 

Yale Grammatical Diversity Project

• Collected attitudes to linguistic constructions from various English varieties 

• Found that people attribute negative stereotypes to people based on their language

Section 3: Why do we do this?

Misconceptions

• There is only one way to use language 

• Languages are immutable

• Language is ‘logical’

Reality

• Every language has different varieties, and language changes over time 

• Language is not logical; they don’t work according to mathematical logic 

  • Language patterns come about through forces of language contact, generalisation, social factors, etc.

Based on the fundamental misconception that certain linguistic constructions have more inherent value than others 

• This is not true! Example: NYC Rhoticity (it used to be a standard variation and had high prestige, but after WWII it became low prestige) 

• Languages change over time, and linguistic constructions have no inherent value

Summary

• Whether a feature of language is evaluated as prestigious or is stigmatised is independent from the linguistic feature itself (associated with the kind of people speaking the variety)

• It is determined by social factors, and it varies depending on time and place

• So, negative reactions to Southern/Female/Black language are a reflection of a negative attitude towards these groups

• Value judgements about language reflect social bias 

Section 4: Linguistic prejudice is detrimental

Example: Vernacular speakers are often misunderstood or unfairly assailed and misjudged in court