What is syntax?

Syntax is a level of grammar that specifically refers to the arrangement of words and
morphemes (the lexicon) in the construction of structures such as phrases, clauses,
and sentences. Syntax can also be seen as the way in which the basically
subconscious rules and categories that are part of each person’s linguistic
competence are used to construct sentences. Syntax deals with the interrelationship
of the elements that make up sentences, and how different rules of arrangement are
used to construct statements, questions, commands, and other types of utterances.

What does it mean to say that syntactic rules are basically subconscious? First,
people apply the rules of their language automatically and without noticing that they
are doing anything special. Second, using the syntax of language is usually obligatory.
Unless you make the grammar explicit (that is, you are consciously aware of it), you
can’t change it. When linguists and anthropologists study syntax, they are interested
in describing the subconscious knowledge that people possess about the syntax of
their language, not prescribing how they should construct sentences. What linguists
and anthropologists are discovering is called descriptive syntax or descriptive
grammar. They listen to what people actually say and then attempt to discover the
rules being used. What a language teacher does in a grammar class by telling
students that there is a correct or incorrect way to write or speak is
called prescriptive syntax or prescriptive grammar.

Syntactic Construction
Sentences are not randomly combined morphemes but structures built on the basis
of rules of combination. The units being combined are called constituents. On page
210 of An Introduction to Linguistics, the authors make an analogy between a
sentence and a car. The largest part is the car itself and can be compared to a
sentence. The smallest parts of the car are individual pieces of metal, rubber, glass,
and plastic and could be compared to individual words. Between the whole and the
individual parts are various assemblies of parts that go together; these are
comparable to phrases. A sentence as made up of small meaningful units (words).
These units combine to make large units, and then these larger units combine into
even larger ones until we have the entire sentence as the largest constituent of itself.
As I noted in video 1, I personally prefer to refer to the clause rather than the
sentence because it is a clause that expresses one complete idea (with a “subject
and predicate”, or “subject and verb”). In contrast, a sentence may consist of one
clause, but it may have two, three or even four clauses which are joined in different
ways. However, many linguists use the sentence as the basis for syntactic
construction. What we can definitely agree on is the existence of identifiable words
and phrases. In order to discuss how syntax can be analysed I will use the following
example, which is both a single clause and a simple sentence:
Artificial Intelligence will become a more important part of everybody’s lives during
the current decade.
 Words
The smallest constituents of a sentence are the morphemes that make it up.
Morphemes make up words. A lexicon for a specific language is a list of all of the
morphemes that are used in that language to form words. The lexicon specifies
whether each morpheme is a prefix, suffix, or root. If it is a root, then the part of
speech (lexical category) can also be included. Here is an example:
Artificial Intelligence will become a more important part of everybody’s lives during
the current decade.
Adjective – Noun - Auxiliary - Verb – Determiner – Adverb – Adjective – Noun –
Preposition – Noun – Noun -Preposition -Article – Adjective
Each word can be labeled, but it is possible that the same word can be used in more
than one lexical category. For example, more can be an adverb or an adjective. The
lexical categories are usually abbreviated as follows:
Adjective – Adj
Adverb – Adv
Noun – N
Verb – V
Auxiliary – Aux
Determiner – Det
Preposition – Prep
Article – Art
Any co-occurrence restrictions are also mentioned by labeling each root as to the
subtype of the category, such as whether a verb is transitive or intransitive. A co-
occurrence restriction is a limitation on the use of a morpheme. For instance, a
transitive verb is limited to sentences that have a direct object.

 *The boy threw.

This sentence is ungrammatical because threw must co-occur with an object as in

 The boy threw the ball.

Phrases
A phrase is any constituent of a clause. Phrases are commonly named for one of
their main elements: noun phrases, verb phrases, adjective phrases, adverb phrases,
and prepositional phrases. Technically, a phrase may be a string of words or just one
word (although in teaching and everyday life we only usually say phrase when there
are two or more words combined). One phrase also can be embedded within
another phrase. In our example there are several phrases:
Artificial Intelligence (intelligence is also a phrase)
will become a more important part (a more important part, important part,
and part are also phrases)
of everybody’s lives (everybody’s lives is also a phrase)
during the current decade. (the current decade is also a phrase)
However, not every combination is a phrase. A more does not function as a
meaningful unit for the sentence. Neither does the current. The current decade
makes sense on its own as well as having meaning in the sentence. The manner in
which constituents of a sentence are arranged is related to the meaning of the
sentence. However, constituents can be arranged ungrammatically and the sentence
might still have the same meaning as it would have had if the constituents had been
arranged grammatically. A native speaker of English will recognize that the following
sentence is ungrammatical, but might still understand what it means:
Intelligence Artificial will become a part more important of lives everybody’s during
the decade current.

The following information about phrases can also be found in chapter 3 of An

Introduction to Linguistics
The head of a phrase is the word that determines the syntactic or phrasal category
of that phrase—whether the phrase functions as a noun phrase, verb phrase,
prepositional phrase, and so on.
The head of a noun phrase is a noun, the head of a verb phrase is a verb, and the
head of a prepositional phrase is a preposition. If the phrase is made up of one word,
then that word is the head of the phrase. If a phrase has two or more words in the
lexical category that the phrase is named for, then the one that carries the central
meaning of the phrase is the head of the phrase. In the noun phrase the boat, it is
clear that boat is the head of the phrase. However, in the noun phrase the title of the
new movie, there are two nouns, title and movie. Because the phrase is about the
title of the movie and not about the movie itself, the head of the phrase is title. All
parts of a phrase that are not the head are called the phrase’s dependents. In some
approaches to syntax, these dependents are further broken down
into specifiers and complements. In the boat, the is the specifier. In the title of the
new movie, the is the specifier and of the new movie is the complement. The
specifier makes the meaning of the head more precise. Determiners are specifiers
for nouns, adverbs are specifiers for verbs, and degree words such
as very and more are used as specifiers
of adjectives and prepositions. Complements provide further information about the
head. The phrase the new movie indicates the title is that of a movie as opposed to a
book, a magazine, or play.
Some languages, including Spanish and English, tend to place complements to the
right of the head (head-first or right-branching languages). Other languages, such as
Turkish, Korean, and Japanese, tend to put complements to the left of the head
(head-last or left-branching languages) with Japanese doing this almost exclusively.
Noun Phrases Among other functions, a noun phrase (often called a nominal phrase)
can function in a sentence as the subject, direct object, and indirect object. A noun
phrase could be a single noun or pronoun or a variety of longer forms:
1. Julian mailed a letter.
(Julian is a noun phrase and the subject of the sentence; a letter is also a noun
phrase and the direct object.)
2. Mary ate the hamburger.
(Mary and the hamburger are noun phrases. Mary is the subject of the sentence; the
hamburger is the direct object.)
3. Three people came late.
(Three people is the noun phrase and the subject of the sentence.)
4. The girl went into the house.
(The girl and the house are noun phrases; the house is an indirect object.)
5. He gave the card to me.
(He is a noun phrase, as is me and the card; me is an indirect object; the card is a
direct object.)

Noun phrases can be abbreviated as NP. A noun might be preceded by an adjective

or adjective phrase. The adjective phrase might include an adverb (very fast horse)
or a subtype of adjective called a determiner (abbreviated as Det).
In English, determiners fall into the following categories: definite and indefinite
articles, demonstratives, possessives, and interrogatives. determiners function to
limit what the noun is referring to, such as to specify whether the referent is a
specific thing or a general thing. Articles (a(n) the), abbreviated as Art, tell whether a
noun refers to a definite (specific) thing, as in the art student, or something that is
not specified (a general thing), as in a very beautiful painting. In the phrase the art
student, the is a definite article. The art student refers to a specific student to
whom we might give a name. In the phrase a very beautiful painting, a is an
indefinite article because the phrase does not specify exactly what painting is being
described. In English, articles are placed before the noun that they modify. They
share this characteristic with demonstratives (this boy), possessives (my car), and
interrogatives (which house).
Some possible noun phrases are as follows:
a. Jim NP N (N is the abbreviation for noun if it cannot be broken down further.)
b. he NP Pro (Pro pronoun)
c. the dog NP Det N (Det determiner, which in this case is an article)
d. six dogs NP Num Noun (Num numeral)
e. the six dogs NP Det Num Noun
f. my dog NP Det N (This determiner is a possessive.)
g. what dog NP Det N (This determiner is an interrogative.)
h. that dog NP Det N (This determiner is a demonstrative.)

Verb Phrases All English sentences (sentence is abbreviated as S) contain a noun

phrase (NP) and a verb phrase (VP); that is, an English sentence is minimally as
follows:
S NP 1 VP
Intransitive verbs can form a verb phrase by themselves. In the simple sentence,
Fish swim
swim is a verb phrase composed of just a verb (VP → V).
Verb phrases often include a noun phrase. Verbs that combine with a noun phrase
are called transitive verbs. In the sentence
Mary ate the hamburger
ate the hamburger is the verb phrase. It can be written as VP → V NP. the
hamburger is the noun phrase within the verb phrase.

Other Types of Phrases In addition to noun phrases and verb phrases, other
important phrasal categories are adjective phrases (AdjP), adverb phrases (AdvP),
and prepositional phrases (PP). Adjective phrases are headed by an adjective but
might also include adjective modifiers (elements that add a property to another
lexical
item). Adjective phrases in turn modify nouns. Adverb phrases are headed by an
adverb and might also include other adverbs and an adjective phrase or phrases.
Adverb phrases modify verbs in the following ways:
1. frequency (They came every day.)
2. duration (The students have been coming for the past five days.)
3. time (Tim will be here at 3 o’clock.)
4. manner (You should do it this way.)
5. purpose (Christopher brought his report card home to show it to his father.)

Prepositional phrases are headed by a preposition and include a noun phrase. Both
adjective and adverb phrases can use prepositions. The question becomes should
the phrase with the preposition be called a prepositional phrase or an adverb or
adjective phrase? Consider the following sentence:
The farmer from Iowa is going into the store.
There are two prepositional phrases in this sentence: from Iowa and into the store.
The function of from Iowa is to modify the farmer; it tells you where he is from. It is
an adjective phrase, but because it is also a prepositional phrase some linguists and
grammar teachers would call it an adjectival prepositional phrase. The phrase into
the store is an adverb phrase (or an adverbial prepositional phrase). It modifies the
verb by telling us where the farmer went.
 Labels and Tree Diagrams
The phrase constituent structure, along with labels of each word, can also be
represented in a tree diagram. Each point at which branching occurs is called a node.
Notice that tree diagrams are upside down. What should be the root is at the top,
and what would be the top is represented by the most specific constituents (the
individual words). The aim is to represent the hierarchical structure of language. That
is, one constituent is often a constituent of a higher level or is a dominant
constituent. All constituents are part of the highest-level or most dominant
constituent, the sentence (=clause) itself.
A clause in English must contain a NP and a VP. You will see in Figure 5-2 that there
are three primary constituents or heads. Auxiliaries are part of the verb phrase, but
they are often marked as a separate head because they are seen as a separate
mental construct. The auxiliary says something about the time frame (that is tense)
of the action or state of the subject noun phrase. However, tense is sometimes
indicated by an inflectional bound morpheme (e.g. with regular past tense verbs,
such as “looked”). In this case, the auxiliary is considered part of the form of the
main verb but Aux is often still put in the phrase marker even though there is no
separate word
Clause (Sentence) Structure
In traditional approaches to grammar, a sentence is seen as having at least two main
constituents; one is called a subject, and the other is called a predicate. In these
traditional approaches the subject is the topic of the sentence and the predicate is a
comment or assertion made about the topic. Everything that is not the subject can
be labelled as the predicate (including words which come before the subject).
In most modern syntactic analyses, the predicate is seen as an element that assigns a
property to another element or elements in a sentence, or helps to relate those
other elements to each other thereby completing the meaning of the predicate.
These expressions are called arguments. Depending on the type of verb, one or more
argument might be needed (obligatory). In addition to arguments that are obligatory,
a predicate may be completed by elements, called adjuncts, which are optional
elements. For instance, in the sentence— The art student looked at a very beautiful
painting—the elements art, very, and beautiful are adjuncts. Adjuncts add
information that is not essential to the meaning of the predicate.
Using a traditional approach, sentences can be classified on the basis of how many
subjects and how many predicates they contain and the types of clauses they
possess. When a sentence consists of only one subject (topic) and one predicate, it is
called a simple sentence. An example of a simple sentence is:

 The dog ran away.

Simple sentences can be combined to form compound sentences, such as:

 The dog ran away and the cat stayed.

 In this case, two sentences are combined using the coordinating conjunction and.
Compound sentences can be formed without a coordinating conjunction, as in the
following sentence:

 We studied all day for the test; now it is time to rest.

In this case, the semicolon takes the place of the conjunction. The two simple
sentences in a compound sentence are said to be independent clauses.
A second type of clause used to construct sentences is the dependent clause. A
dependent clause cannot stand alone as a simple sentence, but must be attached to
an independent clause. A dependent clause often begins with a relative pronoun or a
subordinating conjunction. Some examples of dependent clauses are the following:

 although it is tempting
 who would be traveling with us
 if I come late

A sentence that contains a simple sentence and one or more dependent clauses is
called a complex sentence. The following are complex sentences:

 Although it is tempting, I will not be going to Las Vegas.

 These are the people who would be traveling with us.
 If I come late, start without me.

Sentences that have two or more independent clauses and at least one dependent
clause are called compound-complex sentences. For example:

 When the teacher assigned the reading for the exam, many students were
stunned, but they agreed to study.

When the teacher assigned the reading for the exam is a dependent clause; many
students were stunned could stand alone as a simple sentence and is, therefore, an
independent clause of the larger sentence. The same is true of they agreed to study.
This independent clause is attached to the rest of the sentence by the coordinating
conjunction but.
The terms simple sentence, compound sentence, complex sentence, and compound-
complex sentence refer to the grammatical construction of a sentence. Sentences
can also be classified on the basis of their meaning, purpose (function), or voice. The
following are some of the most common sentence types classified in these ways:

 Declarative—These sentences make a statement. Christine just arrived.

 Interrogative—These sentences ask a question. Has Andrew just arrived?
 Imperative—These sentences express a command or make a request. Aaron,
come here.
 Exclamatory—These sentences show strong or sudden feeling. Oh, if Jan
were only here!
Active or Passive Voice—A voice is the relationship of the grammatical subject of a
verb to the action conveyed by that verb. In most English sentences, the grammatical
subject precedes the verb. In an active sentence, the grammatical subject of the verb
carries out an activity or purpose, as in the sentence Mark hit the ball. In the passive
version of this sentence, the subject is receiving the action of the verb. So in the
above example, what was the direct object becomes the grammatical subject and
what was the grammatical subject is moved to the position of the object. The result
is The ball was hit by Mark. Note that the word by and an auxiliary verb was are
added in this passive construction. Although the word by often indicates a passive
construction, it does not have to be present in a passive sentence. The sentence The
computer was purchased yesterday is also passive. In this sentence the subject (I, we,
a person’s name, etc.) is missing altogether. A possible active version of the sentence
would be I purchased the computer yesterday. In these two examples, the verbs hit
and purchased are in the passive voice.

Phrase Structure Rules

Phrase structure rules specify how constituents of an utterance are arranged and
what constituents can occur as parts of other constituents (the hierarchical structure
of a sentence). A sentence can be described by listing a series of phrase structure
rules starting with the most general (the top of a phrase marker) and ending with the
most specific rules (the bottom of the phrase marker). S → NP AuxVP is an example
of a rule. You can find various examples of rules as applied to sentences in the pdf
study material for this unit.
Lexical categories tend to be spelled out when they can be broken down further,
and abbreviated when not as N in some places and spelled out in others. If we keep
writing rules for individual sentences and then combine them all, we would have a
complete grammar of the syntactic component of English. Such a grammar would be
called a generative grammar: a finite set of rules that could hypothetically produce
an infinite number of utterances. The production of infinite utterances is made
possible by the recursive property of language. Recursion allows one type of
syntactic structure to be included inside another structure of the same type (such as
a noun phrase) to create infinitely long sentences or an infinite number of different
sentences. Other types of sentences (interrogative, compound and complex) can
also be listed.

Noam Chomsky and Generative Grammar

In 1957, Noam Chomsky published a book called Syntactic Structures. He proposed
that humans have a universal innate ability to learn and analyze linguistic
information. This universal grammar provides the general rules
that allow us, at least as children, to learn any language, even with minimum input
from the environment. A child uses it to proceed from the general rules of all
languages to the rules specific to his or her own language. Before this, linguists
tended to emphasize linguistic performance (what the speaker actually says),
whereas Chomsky emphasized linguistic competence, what the speaker
subconsciously knows about his or her language, and what might be called the deep
structure. This led researchers to work to understand how the deep structure is
transformed into the surface structure and how an infinite number of utterances can
be generated from a finite number of rules and lexical items.
In the 1980s, Chomsky introduced the principles and parameters theory which
postulates that even though there is variation in languages, these differences have
specific principles and parameters (limitations). For instance, modifiers, such as
adjectives and adverbs, can come before or after the thing that they modify, but not
several words away from what is being modified. So the language acquisition device
is “programmed” to allow certain possibilities and not others. Children have to learn
from the speaking environment which possibility fits their language. In 1995,
Chomsky further modified his concepts with the formulation of what is called
minimalism. He eliminated the concepts of deep and surface structure as well as
other features of earlier conceptualizations of syntax.

Transformational Rules
Consider the following sets of sentences:
A1 The boy passed out the candy.
A2 The boy passed the candy out.
B1 Linguists often use large words.
B2 Large words are often used by linguists.
Each of the sets above contains synonymous sentences. Yet the forms of sentences 1
and 2 of each set are different. The phrase markers used to represent these
synonymous sentences would be different. Transformational rules (T-rules) relate
the spoken form of sentences to their underlying meaning. More technically,
transformational rules relate the surface structure of sentences to their deep
structure. Different surface structures may have the same deep structure, or
different deep structures may have the same surface structure. We intuitively judge
sentences A1 and A2 to have the same deep structure. Yet the sentences take
slightly different forms. Passed out is an example of certain types of verbal
expressions, called phrasal verbs, that include a verbal base, the main part of the
verb, and a verbal particle (Prt), in this case a preposition. In the sentence, The boy
passed out the candy passed out is the verb, passed is the verbal base, and out is the
verbal particle. Verbal particles are considered to be part of the verb. Yet they can be
separated from the verbal base. Sentences A1 and A2 show that the preposition can
occur on either side of the direct object noun phrase. However, in other sentences
the preposition can be restricted to one side of the direct object noun phrase. In the
following example, the verbal particle is restricted to the left side of the direct object.
C1 Please go over your homework tonight.
C2 *Please go your homework over tonight.
 Of the two sentences listed in A, A1 might be judged to be more basic, since the
verbal base and verbal particle are together. If we take this as our assumption, then
sentence A2 is a transformed version of A1. The rule that relates sentences A1 and
A2 to each other is called the particle movement transformation.

Transformations explain three other processes in addition to movement rules:

deletion, insertion, and substitution.

Deletion Transformations
A sentence that undergoes transformation must have the same meaning as the
sentence from which it was derived. Transformations never change meaning. In the
imperative sentence

 You come here

the pronoun may be deleted. The derived sentence

 Come here

has the same meaning as the basic sentence. This transformation is called the
imperative transformation.

Redundant elements in the deep structure of the basic sentence may also be deleted.
For instance, in the sentence

 If Stephen says he will study for the test, he will study for the test

all but the auxiliary of the second verb phrase can be deleted. The result is

 If Stephen says he will study for the test, he will.

This transformation is called the verb phrase deletion rule.

Insertion Transformations
Words inserted into a basic sentence may not add meaning to the basic sentence. In
the sentences

 A1 He knew she was here

 A2 He knew that she was here

that is inserted in the second sentence. But that has no meaning. In this case, the
addition of that is optional. However, consider the following:

 B1 *He won the race is history.

 B2 That he won the race is history.

Even though the meaning might be clear, B1 is not a grammatical English sentence.
The that in B2 is a word inserted to introduce the noun phrase he won. Although he
won is a grammatical sentence, here it is a part of the larger sentence. In sentence
B2, he won is a NP that is the subject of the sentence. A sentence that is part of
another sentence is called an embedded sentence. So he won is an embedded
sentence, acting as a noun phrase in the larger sentence. The insertion of that to
form a surface structure sentence is, not surprisingly, called the that insertion
transformation.

Substitution Transformations
The only substitution transformations are those that substitute a pronoun for some
other part of speech or syntactic category. For instance

 Anthony thought that Anthony was the best

can become

 Anthony thought that he was the best.

This substitution of a pronoun is called a pronominalization transformation. Like all

transformations, it does not change meaning.

As with phonological rules, transformations can be optional (stylistic) or obligatory.

The particle movement transformation, topicalization rule, imperative
transformation, verb phrase deletion rule, and pronominalization are all optional
rules in English. They may or may not be applied. Yes/no question formation follows
an obligatory rule.

 Will Aaron eat his lunch?

In Japanese, a yes/no question is formed not by a movement transformation as it is

in English, but by an insertion transformation. The suffix –ka is inserted on the end of
a verb to form a question from a statement, but the order of the words is not
changed.

 Kyou (watashi wa) gakkou ni ikimashita. ( (I) went to school today.)

 Kyou gakkou ni ikimashita ka? (Did you go to school today?)

Match the terms and definitions.

Descriptive grammar refers to the mostly subconscious rules of a language that one
uses to combine smaller units into sentences.
 Transformational or generative grammar is a model of syntax that includes a finite
set of rules that could hypothetically produce (generate) an infinite number of
utterances.

A constituent is one unit which is combined with others to create larger syntactic
constructions.

A tree diagram is an illustration that shows the constituents of an utterance, with the
most general at the top and more specific constituents at the bottom.

Phrase structure rules specify how constituents of an utterance are arranged and
what constituents can occur as parts of other constituents.

Grammaticality Judgments
Fluent speakers of a language possess enormous subconscious knowledge, known as
linguistic competence, of the rules of their own language. There is more knowledge
of language in the mind of a fluent speaker than in all the
grammar texts combined. On the other hand, we know nothing about the rules of a
language we do not understand. If you are a fluent speaker of a language, your
subconscious knowledge allows you to produce grammatical or well-
formed sentences. A sentence is grammatical if the sequence of words and the
relationship between words conforms to the syntactic knowledge (rules) of fluent
speakers of a language and if the sentence contains all of its required components.
A fluent speaker also will immediately recognize that certain sentences
are ungrammatical or ill-formed. A sentence is ungrammatical if the sequence of
words and the relationship between words does not conform to the syntactic
knowledge (rules) of native speakers of a language or if the sentence does not
contain all of its required components. You cannot just randomly arrange lexical
items to create a sentence.

 *That man not honest

Any fluent speaker of English would recognize this sentence as being incomplete and
could identify why. The necessity for a verb is one of the more consciously known
rules of English syntax (although not every language requires a verb in every
sentence).

 *Classes the online are.

 English speakers will also recognize that this sentence does not have the correct
English word order (although the order might be correct in another language). Word-
order rules are relative to each language. English usually uses the word order
subject-verb-object (S-V-O). Others may usually have an S-O-V word order, as in
Japanese. Irish uses V-S-O and Fijian uses V-O-S. Languages can be found that use O-
V-S or O-S-V. Linear word order is often specific for the type of sentence. For
instance, the S-V-O word order of English describes declarative sentences, but not
interrogative (question) sentences. In the interrogative sentence

 Did he see a man?

the helping (auxiliary) verb is at the beginning of the sentence.

Case indicates the function of nouns, pronouns, and adjectives within a sentence
and the relationship of those words to verbs and other words within the sentence. In
English and other analytic languages, linear word order alone usually indicates the
grammatical function of a word (its case). In the sentence

 Dogs chase cats

dogs is the subject of the sentence (nominative case) by virtue of its placement
before the verb. If the sentence was

 Cats chase dogs

cats are doing the chasing and the word cats is the subject of the sentence. In
addition to placement in the sentence, the form of the word or an inflectional
morpheme might indicate its case. In English, I, me, and my have the same meaning;
that is, “the person speaking or writing.” The form I indicates that the person is the
subject of the utterance (nominative case); the word me indicates that the person is
the object of the verb or preposition (accusative case), and the word my indicates
the possessive or genitive case. With a word such as Jack’s, as in

 This is Jack’s coat

the insertion of the inflectional bound morpheme -’s indicates the possessive case.
Some languages have many more inflectional bound morphemes to indicate case but
modern English has just two,’s and -s’. In fact, in languages that have many case
endings, linear word order is not always important to indicate case. The inflectional
morpheme alone tells what the function of the word is in the sentence. For example,
the Latin word domus means house. The -us marks the word as being the nominative
case (singular), meaning that it is the subject of the verb. If the word ends in the
bound morpheme –i (domi), it is in the singular genitive case (of the house). An –
o bound morpheme marks the word as being singular and in the dative case (indirect
object). So domo could mean to the house. The morpheme –um, as in domum, marks
the singular accusative case (direct object) So domum would mean house in the
sentence He bought the house. (Each case also has a distinct ending for the plural.)
The word domus would be the subject of the sentence regardless of its position in a
sentence. The word domo would be the indirect object or if it were at the beginning,
middle, or end of the sentence. And the word domum would be the direct object no
matter the word order of the sentence.

Some lexical categories of words can occur together and others cannot. For instance,
“the looked” is not a possible combination of words in English. The articles a, an,
and the do not occur before verbs. Articles occur before many types of nouns or a
gerund. A gerund is a verbal form ending in –ing and it acts as a noun. An example
would be “the running of the bulls in Pamplona.”

Grammaticality of an utterance is not based on whether or not you have heard that
utterance before. Language is productive, so that most of the sentences you create
and hear or see (if written or signed) you have not experienced before. Also, the
grammaticality of an utterance does not depend on whether you understand the
words in the utterance or not. You might not understand the sentence, “Polystyrene
microbeads can be coated with a specified sensing ligand.” However, this sentence is
grammatical and would be understandable to a person who knew the meaning of all
of the words. The opposite is also true. You might understand the meaning of a
sentence that you judge not to be grammatical. Consider the sentence, “The people
is in the room.” This sentence is understandable but ungrammatical. Grammaticality
does not depend on factualness. The sentence, “The president of the United States is
a three-year-old cat” is grammatical, but not factual. The grammaticality of an
utterance is not based on whether or not the utterance makes sense.

What is semantics?
Semantics is the study of meaning in language. A fluent speaker of a language will
know the meanings of many words and sentences. Native speakers of languages are
the primary source of information about meaning. The student (or the professor) of
semantics may well be good at describing meanings, or theorizing about meaning in
general, but he has no advantage over any normal speaker of a language in the
matter of access to the basic data concerning meaning.
English, like most languages, has a number of different dialects. Just as the
pronunciation of English varies from one dialect to another, so there are also
differences in the basic semantic facts from one dialect of English to another.
Meaning is generally studied on two levels: word level (called lexical semantics) and
phrase level (compositional semantics). Some people call the second area “structural”
semantics. Meaning itself can be understood through the terms sense (one mental
representation or concept) and reference (the collection of all the referents of an
expression which exist).

Lexical semantics
 We often ask what a word “means” or we say that a word “means” something else.
What must occur or exist so that we can have this discussion? How are meanings
stored in the mind?
We may be able to recall dictionary-style definitions. But we require words to create
that definition, and those words need to be defined as well. Thus we have an endless
or “circular” process which has no obvious starting point.
We have mental image definitions of many words. However, individual images can
differ. For example, it would not be possible to capture all the referents of food in
one image. We typically form a prototype which does not capture the full scope of
meaning but represents the most typical representation which we can recall. That
said, some words are not easy to “imagine” or “represent.” It is hard to conjure up
an image for “the.”
Finally, we rely on usage based definitions. We know “instinctively” when it is
suitable to use a word to convey a particular meaning or grammatical relationship.
On page 427 of Language Files, there is a helpful example of the word cat. We can
understand how this expression is intended to restrict attention to a certain set of
things in the world (things that are cats).

Semantic Relationships

Words can be semantically related. Pots and pans are both containers for cooking.
There are several terms which we use to identify types of relationships:
Hyponymy
Some words have a similar meaning because they belong to the same segment of a
domain. For instance, the words pink, scarlet, and orange are all more specific terms
for colors that derive from the color red. These words share many of the semantic
properties of the word red so they are referred to as hyponyms of red.
Synonymy
Words that have similar meanings, that share the same semantic properties, are
called synonyms. These are words that sound different but mean the same. When
you paraphrase (restate) a sentence that you have read or heard, you are using
synonyms for some of the original words.

 A (woman or lady) always carries a purse with her.

 Consider how the choice of one or the other affects the meaning of the sentence.
What influences are at work when the speaker chooses one or the other?
The words woman and lady may have the same referent. The same adult, female,
human being may sometimes be referred to as a woman and sometimes as a lady.
These synonyms have the same denotation. Their first definition in the dictionary
would be the same. However, they have different connotations; the shade of
meaning for each word is different. The context in which you would use each word is
different.

 She is a real lady.

 She is a real woman.

These two sentences mean very different things. The first sentence tells us that the
referent is polite, kind, and perhaps elegant and proper. The second sentence
implies that she is strong and determined; it may also have sexual overtones.

Homonymy
Homonyms are words that sound the same, have different meanings, and might be
spelled the same or differently. To, too, and two all sound the same, but each word
means something completely different. Flower and flour, rose (the flower)
and rose (past tense of the verb to rise) are other examples of
homonyms. Polysemous words have more than one meaning. The word school can
be “an institution for learning” or “a grouping of fish.”

Antonymy
Words that have the opposite meaning are called antonyms. They are words that
share many of the same semantic properties, but are opposite in at least one of
them. There are three main kinds of antonyms.
Complementary pairs are antonyms that express a binary relationship in which it is
perceived that there is no middle ground, such as the words male/female. A pair
such as conscious/unconscious demonstrates one way in which complementary pairs
can be formed in English: the use of the prefix un-. Other prefixes that can form
complementary pairs are non- and in-.
The opposite of old is young. But young and old are relative to the speaker’s point of
view. From a child’s point of view, people who are over 30 are old. To a senior citizen,
people who are under 65 are young. So old/young is referred to as a gradable pair.
In fact, old means less young and young means less old. They both have the semantic
property of describing the age of a person or animal. But young refers to an earlier
age and old refers to a later age. How much earlier, or how much later, depends on
the context of the utterance and the point of view of the speaker.
Other examples of gradable pairs are: big/little, high/low. A characteristic of
gradable pairs of antonyms is that they are actually members of a larger set of
related words: Humongous–gigantic–huge–big–large–medium–little–small–tiny–
miniscule
 Relational opposites are antonyms that express a symmetrical relationship between
two words. With the antonym pair parent/child, we can say that Brian is the parent
of Kevin. From this we can infer that Kevin is the child of Brian. In the
pair teach/learn, we can say that John teaches the class. Therefore, The class learns
from John.

Compositional semantics

Sentences make “claims” or “propositions” about the world. A proposition can be

true or false. It has a truth value.

 Luxembourg is the biggest country in the world.

This sentence is definitely false, but what is it that allows me to state that? To
understand a proposition we must understand the sense of all the words. We must
have mental representations of the “content words” and also know how they are
syntactically combined. We must be able to determine the reference of the
proposition, which means to understand what the world would have to be like for
the proposition to be true. These are the truth conditions. In the case of this
example, the truth conditions would be that no other country (recognized as such by
the United Nations) would have a larger territory (measured in squared kilometers)
than Luxembourg. This is my interpretation. Of course, some people might be
thinking about population when they read the word “bigger” because it can also
have that sense. In any case, we may not be able to have perfect information, but we
should still be able to understand the truth conditions to a satisfactory level in order
to derive meaning.

We use these propositions to get a sense of meaning. Propositions can be related.

They can be different types of entailment. They can also be incompatible. There is no
entailment between the sentences below.

 I am drinking a beer.
 I am flying a plane.

We need to use some logic to consider entailment. In the following two examples,
the truth of a guarantees the truth of b. If a is true, b is true.

 A All dogs bark.

 B Sally’s dog barks.
 Entailment can be mutual, but not always, and not in this case. If b is true, this does
not mean that a is true.

Words convey meaning but syntactic structure also has an effect.

 Herman loves Pauline

 Pauline loves Herman

The two sentences have the same words, but they are different propositions and
they are not necessarily both true.

The meaning of a clause is a function of the meanings of the words it contains and
the way these words are syntactically combined. Meanings of multi-word
expressions such as ninety-year-old surfer are therefore also compositional (not
merely lexical).

It is impossible to memorize all sentence meanings, so a person must compute word

meanings and apply mental grammar in order to “figure out” the meaning of novel
sentences. However, compositionality fails when we encounter idiomatic
expressions since the entire phrase has to be stored in the lexicon.

What is pragmatics?
Pragmatics is the study of the ways people use language in actual conversations, or
in context. Pragmaticists study both how context helps to determine whether a
particular utterance is appropriate or inappropriate as well as how changes to
context alter sentences’ meanings.
There are many ways in which context can affect the meaning of an utterance.
Consider a simple sentence such as

 He is there now.

If we don’t know the context, this utterance is difficult to interpret, because it

includes many words that don’t inherently refer to something specific. The meanings
of the words HE, THERE and NOW are always determined by the context in which
they are uttered. We know that he refers to a male and that there refers to a place
and that now refers to a time, but these vague meanings alone don’t give us the
precise information that we need to figure out what would be meant by this
sentence when uttered in some context. Considering this sentence in isolation, we
don’t know whom we are talking about, where he is, or when he is there. To
determine which meaning was intended by the speaker, one would need to know
when the sentence was uttered and what the speaker was talking about. In
pragmatics words such as HE, THERE and NOW are called deictic words.
 Any sentence can take on a particular, novel, and distinct meaning relative to a
particular context. Read the next sentence and reflect on what comes to your mind
for a moment. Afterwards, read the two scenarios. Are either of them at all similar to
your idea?

 Can you stop it?

Scenario 1. A pipe in your bathroom is leaking and water is pouring out. A neighbor is
trying to help, and you anxiously ask this question because you might need to go out
and find a plumber.
Scenario 2. Your child is kicking his sister under the table during lunch. You have
already warned him, and now you are irritated. If he does not stop this time, you are
going to punish him.
There is often a more straightforward meaning to an utterance; this is the one which
the majority of people would assume without knowing the context. However it is
important to recognize that then we are actually assuming one particular kind of
context that affects the meaning of an utterance as much as would any other kind of
context.

Types of Context
An utterance’s context can be broken up into several components.
Linguistic context has to do with what preceded a particular utterance in a discourse.
It refers to what others have said earlier in the conversation.
So, for example, the answer “Yes” means something entirely different when it is an
answer to “Do you like guatita?” than when it is an answer to “Is there an activity to
complete in the forum about this topic?” or “Will you marry me?” The linguistic
context of an utterance tells what speakers are talking about: guatita, homework,
marriage, or whatever. The linguistic context is made up of all of the sentences that
have been uttered in a discourse leading up to the utterance in question.
Situational context allows us to refer to things in the world around us even if they
have not been mentioned before in the discourse.
If you have a conversation about “the president” then it should be clear whether you
are talking about the Ecuadorian president, or another one. A sentence such as
Domenica is very tall has a different meaning if Domenica is five years old, ten years
old or eighteen and part of a basketball team. In the final situation she would have
to be taller than the other players on the team.
Finally, social context includes information about the relationships between the
people who are speaking and what their roles are. In Spanish the social context
determines whether we use tu or usted. In Anglo-Saxon cultures it is much more
common to address a manager by his or her first name than in Ecuadorian culture.
Thus, social context affects language use.
 Felicity: Appropriateness Relative to a Context
Speakers use context to figure out meaning and also whether an utterance is
appropriate in any given setting. We refer to utterances as
being felicitous (appropriate relative to the context) or infelicitous (inappropriate in
some way).

 A: What do you do?

 B: I’m a teacher.

 A: What do you do?

 B: # I have a job. #= infelicity

 A: What do you do?

 B: # I like apples.

What seems to be wrong with the final two conversations? In one, the person
answering the question isn’t providing enough detail. In the other, she doesn’t seem
to give an answer that is at all related to the question. There are many different
reasons why it might be infelicitous to utter a particular sentence in a particular
context; the examples above show only two of these reasons.
An utterance may be called felicitous or infelicitous only relative to a particular
context. It is very easy to think of contexts in which the infelicitous sentences could
be uttered quite acceptably. In other words, felicity is a property of utterances, not a
property of sentences.
In general, the speakers of a language know intuitively whether an utterance is
felicitous or infelicitous, just as they know intuitively whether a sentence is
grammatical or ungrammatical. Also, as with grammaticality, judgments of felicity
may differ from one speaker to another. Nonetheless, there are general guidelines
that utterances must follow in order to be deemed felicitous.

Rules for Conversation

The use of language, like other forms of social behavior, is governed by social rules.
Some of these rules are designed to protect people’s feelings by showing respect or
politeness (e.g., rules governing whether you can use a first name in addressing
someone or must use a title and last name). Even more essential are rules designed
to protect the integrity of our communication: rules that allow our communication
to work.
 Humans are rational creatures so when we communicate, we follow rational “rules”.
They are logical outcomes or “assumptions” rather than conventions. They emerge
naturally within societies and enable effective communication.
The philosopher H. P. Grice (1913– 88) formulated the Cooperative Principle in order
to help explain the relationship between logic and conversation. This principle states
that people intend to be cooperative conversational partners. The speaker and
listener understand that what they say is intended to contribute to the purposes of
the conversation. The exact expectations will differ depending on the particular
context. In a business meeting, one is normally expected to stay focused on a specific
topic, unless it is changed in some approved way. Close friends having a chat in a
café would not have the same strict expectations of appropriate conversational
contributions, although each of the participants would still be expected to behave in
ways that further the purpose of the conversation. The following conversation
sounds strange:

 Lisa: How are you today?

 Nuria: Oh, the next World Cup is in the Americas.
 Lisa: Really? I thought the weather would be warmer.
 Nuria: Well, in my opinion, the rice is from yesterday.

Grice’s Maxims
Grice identified four categories guiding the conversational interactions of both
speakers and hearers. These categories are quality, relevance, quantity, and manner.
Each category contains between one and four maxims, but it is common to talk
about the “maxim of quality”, “maxim of relevance”, etc. We follow these maxims to
ensure that our utterances are felicitous and prevent meaningless discourse. It is just
as important for discourse that the hearer assumes that the speaker is following
them. A conversational partner who constantly assumes that the speaker is lying is
just as uncooperative as a speaker who always lies.
There are two maxims of quality.
•Do not say what you believe to be false.
•Do not say that for which you lack adequate evidence.
In order to follow the first maxim, we must also follow the second. It is only when we
believe we have adequate evidence for some claim that we can have much
confidence that we are not saying something false. Nevertheless, people can differ in
what they think is good evidence for their views. It is also the case that in different
contexts, there are different requirements for how much or what kind of evidence
will qualify as “adequate.” For example, consider a claim like the one made in (2).
(2)The venom of the purple-toothed spider isn’t strong enough to kill people.
If a biologist specializing in human reactions to venomous bites uttered this at a
scientific conference, she would need to have met a certain standard of evidence
before she could felicitously incorporate this utterance into her talk. She would need
some knowledge of the kinds of chemicals in the venom and human reactions to
them; she would also presumably have to know about the history of people who had
suffered purple-toothed spider bites and how they had fared. On the other hand,
consider a person—not a biologist—who had been bitten by a purple-toothed spider:
as a result, he got a painful swelling at the location of the bite but was otherwise
unaffected. In chatting with his friends, he might legitimately be able to utter (2)
without knowing anything more general about these spider bites; his evidence
would be only his personal experience. Thus these two individuals speaking in
different contexts have two distinct standards for quality of evidence. Of course, the
second individual might be wrong: it might be the case that he was merely very lucky
and didn’t get very much venom in his body, but a worse bite (or perhaps a bite to a
smaller or less healthy person) could cause death. Nonetheless, he has followed
Grice’s maxims by saying what he does not believe to be false and something for
which he has adequate evidence based on the situation. If someone asked him, “Are
you sure?” he might then consider explaining his evidence or weakening his claim:
something like (3).
(3)Well, when I was bitten by a purple-toothed spider, I didn’t die. So at least I know
that the venom doesn’t always kill people.
Meanwhile, the biologist could likely answer, “Yes; I’m sure,” without further
qualifications (though at a talk she would be expected to be able to produce
evidence for her claim). Even though their levels of certainty differ, both of them
would have equal claim to utter (2) given the appropriate context and their stated
experience.
Students need to understand this issue about standards for quality of evidence when
they write academic papers and have to cite their sources.
b. The maxim of relevance (also called the maxim of relation) is often perceived as
being the most obvious. It is also the most simply stated.
•Be relevant.
This maxim has a central role in maintaining the organization of conversation by
preventing random topic shifts like those found in (1). To avoid such discourse, we
are expected to make contributions that pertain to the subject of the conversation. If
someone asks you about your plans for dinner, you should give an answer about that
topic rather than telling a story about a series you are watching on Netflix.
From the hearer’s perspective, the maxim of relevance helps us to figure out what
others mean by their utterances. Our default assumption is that the people we are
talking with are cooperative and that they are doing their best to make the
conversation work. This assumption allows us to make inferences. Consider the
following conversation:
(4) Hannah: Is Carlos working these days?
Mark: Well, he isn’t living in Guayaquil.
If she did not have these assumptions as a part of her linguistic competence, Hannah
could take Sam’s response to be completely unhelpful. However, Hannah will
assume that Sam intends his contribution to be relevant, so she will likely draw the
inference that Mark doesn’t know if Carlos is working or not. She may also assume
that Carlos has moved away to look for work in another place
As with the first maxim of quality, the maxim of relevance seems perfectly obvious,
but that doesn’t mean that people can’t ever change topics. Imagine that two
roommates have just arrived back in their dorm on a Friday afternoon; the following
is an excerpt from their conversation:
(5) Rachel: We should think of something fun to do this weekend!
Sarah: Can we talk about something else? I want your advice about something.
In (5), Sarah uses the word “else” to show Rachel that she knows she is supposed to
stay on topic and be relevant by discussing weekend plans, but she has something
else on her mind, and she asks for permission to go against that maxim. Of course,
people don’t always point out when they are about to say something irrelevant. We
have all had conversations in which we are trying to discuss some particular topic,
only to have our conversational partner jump in with an unrelated fact or story. We
may or may not be bothered—sometimes we do allow others to go off on
tangents—but we are justified in objecting. Someone saying “Wait a minute! You’re
changing the subject!” acknowledges this jointly held assumption that utterances
should be relevant.
c. The maxims of quantity concern how much information it is appropriate for a
speaker to give in a discourse. Of course, there are some situations in which more
information is needed and others in which less is needed. Notice how the two
maxims of quantity are phrased in order to make allowances for these differences.
•Make your contribution as informative as is required.
•Do not make your contribution more informative than is required.
The first of these maxims reflects the fact that we are expected to give all of the
information necessary for a given circumstance and to make as strong a claim as is
warranted (see the second maxim of quality). The second reflects the expectation
that we neither provide too much information nor make a stronger claim than is
warranted.
Some examples will help to illustrate.
Suppose that you are asked what you are going to do over the weekend. If a
colleague asks you at the end of a meeting, it will likely be acceptable to mention
only one or two things that you intend to do (and it would be both infelicitous and
rude to provide a complete schedule of everything you plan to do). However, if a
friends is trying to arrange to meet up with you, she likely needs to know specific
times that you will be available. In this case, if you were to respond with the same
short answer, it would be under-informative and therefore infelicitous.
Suppose that someone asks you where you grew up. Any of the possible responses
given in (6) could be true answers to the question (and thereby follow Grice’s
maxims of quality and relevance), but it is obvious that some of these answers would
be appropriate in certain contexts and not in others. Each response could be too
informative, not informative enough, or just right, depending on the circumstances.
Try to think of an example of each kind of context.
(6) a. On the corner of Main Street and Minor Road
b. In Dayton
c. In Dayton, Ohio
d. In Dayton, Ohio, on the corner of Main Street and Minor Road
e. In Ohio
f. In the Midwest
g. In the United States

d. The maxims of manner differ critically from the other three sets of maxims. The
maxims of quality, relevance, and quantity all have to do with the information that a
speaker is expected to give or not give in a discourse. The maxims of manner, on the
other hand, have nothing to do with the information itself; rather, these maxims
have to do with expectations about how one goes about giving and interpreting that
information in being a cooperative conversational partner.
•Avoid obscurity of expression. (That is, don’t use words or phrases that are hard to
understand.)
•Avoid ambiguity.
•Be brief.
•Be orderly.
The first maxim, “Avoid obscurity of expression,” indicates that speakers should
avoid the use of jargon (terms restricted primarily to specialized areas of knowledge)
or other terms that their listeners cannot reasonably be expected to know and that
they should also avoid needlessly complex sentence structures.
The second maxim references the understanding that what we are saying should be
clear within the particular context. Speakers should avoid saying things that have
more than one meaning (e.g., He promised to phone at noon: what happened at
noon—the promise or the phone call?) unless their listeners can be expected to
know which meaning was intended. While there are many words and phrases that
would be ambiguous out of context, the listener is expected to interpret the
meaning based on the context. If Polly tells her friend that she’s going to the bank to
deposit a check and the friend claims to be confused as to whether a financial
institution or a riverbank is being referred to, it is the friend (the hearer), rather than
Polly, who is not following this maxim.
The third maxim, “Be brief,” tells us not to say too much when a few words are
sufficient. The expectation to be brief is different from the expectation to not give
too much information. Notice that the speakers in both (7B) and (8B) give exactly the
same amount of information, but they do so in different words. Thus, the speaker in
(8B) violates a maxim of manner because he is being wordy, but he does not violate
a maxim of quantity.
(7) A: What do you do?
B: I’m an English teacher in a high school.
(8) A: What do you do?
B: # What I do is that I’m a teacher in a school for students aged between 12 and 18
and the subject that I teach is English.
The fourth maxim, “Be orderly,” refers to the expectation that what we say should
be organized in some intelligent way. So, if you have information to convey about
several different topics, you should convey all of the information on one topic first,
followed by the next, rather than giving one sentence about each in alternation.
Often speakers follow this maxim by giving general overview information first and
then moving on to specifics.
Telling a story in chronological order also is part of following this maxim. For
example, consider the strangeness of (9) and (10). The first merely sounds peculiar,
while the second is actually hard to follow.
(9) # Leslie read fifty pages and opened her book. My mother didn’t really want my
room to be painted purple. I was worried that I wouldn’t get good grades at the new
school.
(10) # When I was a child, my favorite color was purple. I worked very hard in all of
my classes to get good grades. My mother told me that if I got good grades, I could
paint my room. When I was ten years old, I switched to a new school. I wanted to
paint my bedroom a bright color.
To sum up this section, although the four maxims of manner do not provide any
insight into what information a speaker should share, they are critical with regard to
how that information can be clearly understood.
Flouting Maxims
So far, for the most part, we have considered cases in which speakers follow Grice’s
maxims. Of course, people sometimes violate the maxims: at some point everyone
has told a lie, changed the subject, given too much information, or said something
confusing. Sometimes people violate the maxims on purpose (e.g., lying in order to
intentionally deceive someone), and other times by accident. Strictly speaking, these
violations are infelicitous.
But the maxims can also be exploited or flouted in order to communicate indirectly.
A speaker flouts a maxim when he says something that in its most literal meaning
appears to violate a maxim, but the listener is expected to understand the meaning
being conveyed due to the shared understanding of the maxims. There are several
reasons that one might choose to use the maxims in this way. We sometimes need
to avoid saying something directly because doing so could hurt us or someone else.
Grice gave an example of a professor who was asked to write a letter of
recommendation for a recent PhD graduate who was applying for a teaching position.
Suppose that the letter went like this:
(11)
Dear Colleague:
Mr. John J. Jones has asked me to write a letter on his behalf. Let me say that Mr.
Jones is unfailingly polite, is neatly dressed at all times, and is always on time for his
classes.
Sincerely yours,
Harry H. Homer
Do you think Mr. Jones would get the job? Probably not! In this case, the maxim of
quantity is being flouted. Professor Homer wanted to convey his negative impression
of the candidate without actually saying anything negative about him. The fact that
he gives much less information than would normally be expected for this type of
letter communicates this message clearly. The recipient of this letter will assume
that Professor Homer is intending to be cooperative; the shortness of the letter
indicates that he has said all of the relevant positive things he could think of.
The other maxims can also be flouted. For example, if you and a classmate are
discussing your professor, and you see your professor rapidly approaching, you may
suddenly change the subject by looking pointedly at your classmate and saying, “Oh,
really? I didn’t know that chocolate originated in Mexico!” In this case, you probably
don’t want to change the subject to the history of chocolate; rather, you are hoping
to prevent your classmate from saying anything inappropriate! You expect that she
will notice your abrupt change in subject and deduce that something is up. You have
successfully flouted the maxim of relevance.
It is also possible to flout the maxim of quality. If someone says to you something
that you don’t believe, you may respond, “Right, and I’m the Prince of Persia.” You
don’t mean that you are a prince, rather, you mean something like ‘What you just
said is as obviously false as the idea that I’m the Prince of Persia.’ A sarcastic
comment such as this may sound harsh, but it may be perceived (in some contexts)
as less hurtful than coming right out and saying, “You’re wrong.” Flouting the maxim
of quality can also allow us to insult people and (usually) get away with it. If your
friend is bragging about something mundane, you might say, “That’s the most
amazing thing I’ve ever heard—please tell me more!” In this case your friend will
probably take it as an insult, but not get deeply offended. The insulter is flouting the
first maxim of quality—the recognition that the claim is too strong (see the maxims
of quantity) for it to likely be true.
The flouting of maxims often plays a particularly large role in humor, but it is an
important part of everyday communication; it allows us to draw conclusions and can
facilitate efficient communication. It is important to remember that when speakers
and hearers flout maxims, their intention is to be cooperative conversational
partners. But this cooperation requires a higher degree of effort on the part of both
speaker and hearer, so it carries the risk of the intended message not getting
through. You can probably think of a situation in which you meant an utterance to
be a joke, and the person you were speaking to didn’t “get it” for whatever reason;
this can particularly be a risk when conversing with someone you don’t know well, or
a child, or someone from a different area or culture who does not share your specific
expectations about communication.
Grice’s Maxims in a Wider Context
The needs of social harmony, politeness, and linguistic integrity are not always
consistent with each other. We have already seen several cases in which politeness
keeps us from following pragmatic rules. The rules for conversation are social rules
but sometimes they lose out to other social rules. Grice’s maxims have a general
rationality, but the exact way that they are implemented and the way that they
interact with other societal rules varies between societies.

Drawing Conclusions
A crucial part of understanding utterances is being able to draw conclusions from
those utterances about the way the world is. However, the conclusions we draw can
be based on different kinds of evidence or reasoning. The sorts of reasoning that we
use depend largely on the context of the utterance that we are interpreting.
Drawing Conclusions: Entailment
One kind of reasoning commonly used to draw conclusions is based on the concept
of entailment- For any two sentences X and Y, sentence X entails sentence Y if
whenever X is true, Y must be true as well.
1 In the example in (1), the X sentence entails the Y sentence.
(1) X: Ian eats a large breakfast every day.
Y: Ian eats a large breakfast on Mondays.
Entailment indicates a commitment from the speaker’s point of view. Entailment
also does something for the hearer: from the hearer’s point of view, entailment
allows a conclusion to be drawn very confidently. If you hear and believe X, and X
entails Y, then concluding Y is completely safe. Entailment is a relationship based on
literal meaning. Thus, entailments are conclusions that can be drawn irrespective of
an utterance’s context. But often, if you take only what is literally asserted and
entailed by an utterance, that part of the meaning alone is not enough to account for
hearers’ understanding of the utterance. Speakers routinely intend to convey
information in addition to what is entailed by the sentences they utter. Fortunately,
hearers also routinely draw conclusions from the utterances they hear, even when
the sentence uttered does not entail the conclusion drawn. That is, an utterance’s
context often helps us to draw conclusions—inferences —that were not entailed by
the sentence that was spoken.
Drawing Conclusions: Inference, Implication, and Implicature
People commonly draw inferences from what others say based on the assumption
that speakers are adhering to the Cooperative Principle. First, consider a situation in
which an inference is drawn that does not involve linguistic communication. A
meeting between a supervisor and an employee is running longer than the allotted
time. The employee doesn’t want to say, “Our meeting is running longer than we’d
scheduled,” because the supervisor might find it rude. Instead, the employee glances
at his watch. The employee is implying that the meeting is running long: sending the
message without saying it directly. The supervisor, if he understands the message,
infers that the employee wishes the meeting to end. An inference is a conclusion
that a person is reasonably entitled to draw based on a set of circumstances.
A person may draw an inference in cases when no one has tried to imply anything at
all. If you walk outside and notice that the pavement is wet, you might infer that it
had been raining, but you wouldn’t want to say that the pavement had implied
anything. (There must be someone trying to communicate an idea in order to say
that any implying has happened.) In the rest of this file, however, we will be
considering only inferences drawn when there is a person trying to send a message,
and more specifically, we will consider only cases in which—unlike those above—the
message is sent using language. When a speaker implies something using language,
we say that her utterance contains an implicature. Implicatures are conclusions that
are drawn about what people mean based on what we know about how
conversation works. There are many different kinds of implicature, and we will
consider only a few of them here, namely, those that arise via one of Grice’s maxims
for cooperative conversation.
Implicature Based on the Maxim of Relevance
Consider the following sample of discourse between two strangers at a bus stop:
(2) Speaker 1: I’d really like a cup of coffee.
Speaker 2: There’s a place around the corner called Joe’s.
Here’s a reasonable conclusion Y that we can draw from Speaker 2’s utterance of X:
(3) X: There’s a place around the corner called Joe’s.
Y: Joe’s sells coffee.
It is important to recognize that in (3), X does not entail Y: it is obviously possible for
there to be a place around the corner called Joe’s that doesn’t sell coffee. Thus, the
conclusion of Y is an inference: it is based on an implicature rather than an
entailment.
How does the implicature arise? Speaker 1 is talking about coffee and looking for
information about coffee. If Joe’s were a bookstore that didn’t serve coffee, then
Speaker 2 would be changing the subject, which people usually don’t do in the
middle of a conversation. Speaker 1 is much more likely to assume that Speaker 2 is
following Grice’s maxim of relevance: if he wants to interpret Speaker 2’s
contribution as relevant, he has to “read something into it” that Speaker 2’s
utterance didn’t entail, namely, that Joe’s sells coffee. In order to justify conclusion Y,
we had to think about pragmatic concepts: people and conversation in context. We
say that X implicates Y in this situation.
(4) Hannah: Is Carlos working these days?
Sam: Well, he has moved away from Guayaquil.
The implicature from Sam’s utterance (again based on the assumption that his
contribution is relevant) is that Carlos could not find a job in Guayaquil. Sam might
instead have said I don’t know because he has moved away, but the fact that
Guayaquil is his hometown and he moved away makes me think that he could not
find a job here. Given our set of maxims, though, Sam can say what he does and rely
on the listener to figure out what he means without explicitly stating these other
steps.
It is important to note that if Sam knows that Carlos is not working in the place
where he is living, then his response is either very misleading (if he understands that
his utterance will generate an implicature) or at least infelicitous (if he merely thinks
he is saying something unrelated to the topic at hand).
Implicature Based on the Maxim of Quantity
The conversation in (5) illustrates an implicature that might arise on the assumption
that the speaker is obeying the first maxim of quantity: a speaker should give as
much information as required.
(5) Mother: Have you done your homework for all of your classes yet?
Son: I’ve finished my history homework.
Let us again consider the actual content of what is uttered compared with the
conclusion that is likely to be drawn, shown in (6X) and (6Y), respectively.
(6) X: I’ve finished my history homework.
Y: I have not finished my homework for my other classes.
Clearly, in this case X does not entail Y. It is very possible for a child to say truthfully
that he has finished his history homework and to have also finished the work for his
other classes. Rather, the mother is likely to infer Y because her question wasn’t
looking for information merely about the history homework but rather for
information about work for all of her son’s classes. She will assume that her son is
giving as much of the information as possible that is required to give a complete
answer to her question. Numbers are a particularly common source for the
generation of quantity implicatures. Consider the following discourse. What seems
to be wrong with it?
(7) Gail: How far can you run without stopping?
Kim: Ten miles.
Gail: I guess you can’t run a whole marathon without stopping, then.
Kim: Nonsense, I’ve done it a number of times.
Notice that what Kim says first must be true if what she says next is true. Certainly, if
Kim can run over twenty-six miles without stopping, then she can run ten miles
without stopping. However, Gail quite naturally assumed that Kim was obeying the
first maxim of quantity with her answer of “ten miles”; Gail therefore inferred that
Kim meant ‘exactly ten miles, and no more.’ If you pay attention, you are likely to be
surprised by how often numbers such as 47 are used to implicate ‘exactly 47’ when
the entailed meaning is ‘at least 47.’ These implicatures are so strong in English that
people often view statements like Kim’s as lies, even though what Kim says is
technically true (following the maxim of quality). The deceptiveness of this
statement comes from her clear violation of the maxim of quantity.
Implicature Based on the Maxim of Manner
Recall that one of Grice’s maxims of manner tells speakers to be orderly. Keeping
this in mind, consider the two stories told in (8) and (9).
(8)Rebecca took the medication and had an allergic reaction.
(9)Rebecca had an allergic reaction and took the medication.
Both of these sentences provide exactly the same entailed meaning. However,
someone who assumes that the speaker is being cooperative will assume that the
speaker is telling the story in an orderly fashion. Thus, someone who hears (8) may
infer that Rebecca had an allergic reaction to the medication, whereas someone who
hears (9) is more likely to infer that Rebecca took the medication in order to counter
her allergic reaction to something else. Another one of the maxims of manner
dictates that speakers be brief. Consider the following utterance:
(10)The man who lives with me is an electrician.
Upon hearing this sentence uttered by a person whom you don’t know particularly
well, you might infer that the speaker is talking about a house mate (or an
apartment mate, or something similar). Of course, as far as entailment is concerned,
the speaker could be talking about a husband, son, or brother—all of which might
explain their living together—but because “my husband” is shorter than “the man
who lives with me,” it is likely that the speaker would have used the shorter phrase,
were it true.
Thus, by using the lengthier expression, the speaker implicates that she does not
have one of these other more specific kinds of relationships to the electrician.
Implicature Based on the Maxim of Quality
The second maxim of quality tells us that we can felicitously say only that for which
we have adequate evidence. People often differ in what they think is sufficient
evidence for their views. Sometimes, we may draw inferences based on the
assumption that we have the same standards for evidence as do our conversational
partners. Consider the following conversation:
(11) Sandy: We need someone to make some sort of cake for the picnic.
Tom: I can make my family’s favorite chocolate cake.
Sandy might draw the inference that Tom has made his family’s favorite chocolate
cake before, because the best evidence that Tom can make this cake would be that
he had indeed made it, as spelled out in (12).
(12) X: I can make my family’s favorite chocolate cake.
Y: I have succeeded in making this cake before.
However, this inference is not entailed by Tom’s statement; it is only implicated.
Tom could legitimately say that he could make the chocolate cake based on the fact
that he had a recipe and had watched it being made many times and thought he
knew all he needed to know to make it. Suppose Tom were to make the cake and it
turned out very badly. Something like the following conversation might take place:
(13) Sandy: I thought you said you could make this cake!
Tom: Well, I thought I could.
As Sandy’s challenge—which sounds quite felicitous—illustrates, she is justified in
being upset that Tom did not have a high enough standard of evidence for saying
that he could make the cake. Thus, the inference that she drew was well-founded.
Was Tom justified in saying that he could make the cake in the first place? This
question is one whose answer will be open to differences of opinion. The point,
though, is that we ought to be aware that people may often infer a stronger claim
than what has been entailed, based on their assumption about the sort of evidence
that might be required in order to felicitously express some proposition.
The Significance of Implicatures to Communication
The system of implicature is a kind of side effect of Grice’s maxims. Implicatures
allow us to introduce ideas into a discourse with less commitment than we would
have to express were we entailing the same propositions. In (4), why would Sam
choose to give the answer that he gave instead of saying, “I don’t know,” or
something similar? Whatever his reason, it is clear that he wants Hannah to draw her
own conclusions. Maybe he isn’t certain and doesn’t want to commit. Perhaps he
wishes to be discreet. Implicature gives him a way to communicate the idea he has in
mind while still protecting himself from committing to the truth of a proposition that
he does not want to commit to.
On the other hand, implicature can serve a function much more fundamental to our
conversations than merely protecting noncommittal speakers. One major reason for
exploiting the maxims in this way is to make conversation easier. If we were forced
to speak telling the complete truth in an entirely logical way, making sure that what
we said entailed every fact that we wanted our hearers to conclude, conversation
would proceed very slowly and probably be very monotonous. We use context and
our knowledge about the universe to draw inferences from what we hear because it
allows us to use language more effectively.

Speech Acts
An Introduction to Speech Acts
People perform a kind of act simply by using language; these are called speech acts.
These acts are often called functions in the field of language learning. For example,
we use language to convey and request information, give orders, make requests and
threats, give warnings and advice, etc. Read the following sentences and consider
the function.
(1)Julio was at the office yesterday until 6 P.M.
(2)Who ate all the bread?
(3)Sit down and be quiet.
(4)Please let me know if you’ll be attending.
(5)If you do that again, I’ll report you.
(6)Watch out—there’s a huge pothole in the road there.
(7)I bet you $5 that you get a better grade than me.
(8)You need to respond to some of the questions in the forum after class.
With language we can do things that would otherwise be impossible. In (6), we could
warn someone of a pothole by pointing at it, but only if we were in a position to see
it. How could we give the advice in (8) without words? It would certainly be difficult.
The following list contains some of the most common speech acts. Of course,
language can be used for all sorts of purposes other than those listed, as well.
Some common speech acts and their functions
Speech Act Function
assertion conveys information
question elicits information
request (more or less politely) elicits action or information
order demands action
promise commits the speaker to an action
threat commits the speaker to an action that the hearer does not
want

Felicity Conditions
In order to be felicitous a speech act must be uttered in a certain kind of context. It
must be directed to a person (or animal or machine) that is capable of doing
whatever action was requested. Here are some examples of additional felicity
conditions for two very common speech acts: requests and questions.
Felicity conditions for requests
In order for a speaker to felicitously request a hearer to complete some action, it
should be the case that . . .
a. The speaker believes that the action has not yet been done.
b. The speaker wants the action to be done (or thinks that the action should be done
for some reason).
c. The speaker believes that the hearer is able to do the action.
d. The speaker believes that the hearer may be willing to do things of that sort for
the speaker.
Felicity conditions for questions
In order for a speaker to felicitously question a hearer about some state of affairs, it
should be the case that . . .
a. The speaker does not know some piece of information about some state of affairs.
b. The speaker wants to know that information about the state of affairs.
c. The speaker believes that the hearer may be able to supply the information about
the state of affairs that the speaker wants.
To understand when it is appropriate to make a request or to ask a question, then,
we need to think about the felicity conditions associated with each of these speech
acts. This applies for all other speech acts as well. Some conditions for a speech act
may be suspended in certain contexts such as when we are playing a game, or doing
a role play in an English class. We recognize these situations to be socially
exceptional in one way or another. So, we can modify the particular felicity
conditions but we have to be careful: for example, we wouldn’t want to say that in
the case of a teacher asking a question there were no felicity conditions at all; rather,
there would be a modified set of felicity conditions including perhaps such items as
‘The speaker wants to know whether the hearer is able to supply an answer.’
Performative Speech Acts and Verbs
Any time that you open your mouth and utter a sentence, you perform a speech act.
A special kind of speech act, known as a performative speech act, is one in which the
particular action named by the verb is accomplished in the performance of the
speech act itself. For example, someone can say “I am throwing a ball” without a ball
actually being thrown (the throwing action is separate from an assertion about such
an action), but someone cannot normally say “I promise to take you to the store
later” without actually making such a promise. Performative verbs therefore denote
purely linguistic actions. Compare (12)–(19) with (1)–(8).
(12)I assert that Julio was at the office yesterday until 6 P.M.
(13)I ask again: Who ate all the bread?
(14)I order you to sit down and be quiet.
(15)I request that you please let me know if you’ll be attending.
(16)Yes, I’m threatening you: if you do that again, I’ll report you.
(17)I’m warning you: there’s a huge pothole you need to watch out for.
(18) I bet you $5 that you get a better grade than me.
(19)I advise you to respond to some of the questions in the forum after class.
Most of the time we don’t need to actually utter the performative verb, but some
ceremonies or formal actions require this. A good example is a marriage ceremony:
(20)I hereby pronounce you husband and wife.
In this example the speaker not only performs a speech act but also changes
something about the world: the marriage between two people. However, only a
person who is “authorized” can change something when he or she utters the words.
This is another felicity condition.
If you yell “I quit!” in the classroom in a moment of frustration after a difficult class,
you would not be expected to begin a new job search the next morning as you would
be if you said it to your boss during a staff meeting.
Not all speech acts containing verbs that can be used performatively are
performative speech acts. Consider the following sentences:
(25)I promise I will help you with your project this week.
(26)John promises he will help you with your project this week.
(27)I will promise to help you with your project this week.
Although all of these sentences use the verb promise, only (25) uses it as a
performative verb. Sentence (26) is an assertion about someone else’s promise, and
(27) is an assertion about a future promise the speaker will make, so neither of these
is a performative speech act. Why? There are two major requirements for
performatives: (i) the subject of the sentence must be first person, I or we, since
these speech acts concern the interaction between speakers and hearers; and (ii) the
verb must be in the present tense, since performative speech acts, like all actions,
take place in the present. Sentences (26) and (27) are therefore not promises
because the subject of the sentence is third-person John, and the verb is in the
future tense, respectively.
Direct and Indirect Speech Acts
The types of speech acts that we have been considering, including both performative
speech acts and the examples in (1)–(8), are called direct speech acts, because they
perform their functions in a direct and literal manner. That is, the function that the
sentence performs in a discourse is evident from its literal meaning. However, we
very commonly perform speech acts indirectly, especially when we are trying to be
polite.
(31) Questions
A. Direct
a. Did John marry Helen?
b. I’m asking you whether John married Helen.
B. Indirect
a. I don’t know if John married Helen.
b. I would like to know if John married Helen.
c. Do you know whether John married Helen?
(32) Requests
A. Direct
a. (Please) Take out the garbage.
b. I request that you take out the garbage.
B. Indirect
a. The garbage hasn’t been taken out yet.
b. I would like for you to take out the garbage.
c. Could you take out the garbage?
d. Would you mind taking out the garbage?

Indirect speech acts have a very close connection with the felicity conditions on
speech acts; we can perform an indirect speech act in many cases by appealing to a
particular one of its felicity conditions. At the same time they are often, although not
always, indicative of politeness considerations on behalf of the speaker. Instead of
assuming that the listener is prepared to take out the garbage, the speaker might ask,
Would you mind taking out the garbage? in order to make a polite request.

Identifying Indirect Speech Acts

In an indirect speech act, what the speaker actually means is different from what she
or he literally says. There are several ways to determine whether an utterance is an
indirect speech act. First check to see whether it is a performative speech act, since
those are always direct. For example, (31A.b) and (32A.b) both contain performative
verbs, and therefore both perform direct speech acts. If the speech act is not
performative, it might be indirect.
We can also check to see whether any felicity conditions are violated for the
sentence’s literal meaning but not for its intended meaning. If any are, then the
sentence must be an indirect speech act. For example, if taken literally, (32B.c)
would be a question asking whether the hearer is able to take out the garbage.
However, if the listener understands that the speaker already knows that he is
capable of the action, this sentence is not a direct speech act of questioning, but an
indirect speech act of making a request.
Different speech acts arouse different responses. Listeners respond to an assertion
by a signal of acknowledgment, such as a nod or a verbal response like Oh, I see.
People respond to a question by a confirmation or denial or by supplying the
information being solicited. People respond to a request or command by either
carrying out the action accordingly or refusing with some explanation.
If the standard response to an utterance is different from what its literal meaning
would arouse, then it is used to perform an indirect speech act. For example, as
noted above, the literal interpretation of (32B.c) would be a question. But compare it
with something like Could you lift 200 pounds? You can respond with a simple Yes, I
could or No, I couldn’t, but it is not appropriate, felicitous, or polite to respond to
(32B.c) with only this. Instead, people normally respond to such an utterance by
actually carrying out the requested action—taking out the garbage, in this case. This
shows that while Could you lift 200 pounds? is usually a direct speech act of
questioning, (32B.c) is usually an indirect speech act of requesting: it has the same
effect as (32A.a).
Sentences and Their Relation to Speech Acts
There are many different ways to perform the same speech act, because there are
many different sentences that will accomplish the same goal. We have the choice
between speaking directly (with or without performatives) or indirectly. We can also
choose a particular sentence type.
Certain speech acts are so common that many languages have particular syntactic
structures conventionally used to mark them. Some examples of different types of
sentence structures for English are given in (33), along with a basic notation of the
order of subject (S), verb (V), and object (O) as a shorthand for their major syntactic
characteristics.
(33) Sentence Type Examples
Declarative He is cooking the chicken.
Interrogative Is he cooking the chicken?
Who is cooking the chicken?
What is he cooking?
Imperative Cook the chicken.
On the surface, it looks as though declarative sentences, which in English usually
follow the basic word order of SV(O), are perfect for making assertions. Interrogative
sentences, which usually have a verb form and/or a wh- word like who or what at
the beginning of the sentence, are designed for asking questions; and imperative
sentences, which usually lack a subject (sometimes referred to as “understood you”)
are made for giving orders. This association is fairly typical and often holds. But don’t
confuse the sentence types (declarative, interrogative, and imperative) with speech
acts (assertion, question, and request)! They are different, and this association does
not always hold.As with all things related to pragmatics, the key is context.

Presupposition
Presuppositions of Existence
(1)The Amazon River runs through northern Europe.
Most readers probably have a fairly strong reaction to sentence (1). Some of you
may have thought, “Oh, yeah, I knew that,” but presumably most of you responded
by thinking something like, “No it doesn’t! The Amazon River is in South America!”
Regardless of which reaction you had, however, none of you responded by thinking,
“But there’s no such thing as the Amazon River,” or wondering whether there’s a
place called Europe. If you were having a conversation with someone who asserted
(1), you would do very well to disagree with that person, but you would be
disagreeing about the location of the Amazon River, not its existence. Compare your
reaction to (1) with the sort of reaction you might have to (2).
(2)The Bvryzax River runs through northern Europe.
Could you respond to (2) by saying, “No it doesn’t!” Indeed, you could not. Why?
Because (at least at the time of this publication) there is no river anywhere in the
known universe by the name of Bvryzax. In order to say of a river that it does not run
through northern Europe, you must believe that the river exists. If you wanted—very
rightly—to object to someone’s uttering (2), you would have to say something more
along the lines of “There’s no such thing as the Bvryzax River.” Similarly, if a child you
know tells you that the monster under her bed has fangs, you would likely not want
to say, “No, it doesn’t.” Responding in that way would merely corroborate the
existence of the (perhaps fanged) monster. Rather, you would want to dispute the
child’s underlying assumption that a monster existed at all. Both (1) and (2) would be
infelicitous in almost any context that you can think of—other than perhaps a work
of fiction—but they would be infelicitous for different reasons. An utterance of (1)
would be infelicitous because of a violation of Grice’s maxim of quality. An utterance
of (2) would be infelicitous because it presupposes the existence of something that
in fact does not exist. A presupposition is an underlying assumption that must be
satisfied in order for an utterance to make sense or for it to be debatable.
Presuppositions appear exceedingly often in the sentences that we hear uttered
every day, and most of the time we don’t notice their presence at all. However,
when they are not satisfied, we are often left not knowing quite how to respond.