University of Birmingham School of Computer Science

A Glossary of Linguistic Terms

Dr Peter Coxhead

Warning: This web page was originally constructed to help computer science students who were taking my module on natural language processing. Some terms may be used differently by different authors. Unless otherwise stated, definitions are based on the English language.

If you find any errors, please e-mail me at p.coxhead@cs.bham.ac.uk.

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

A

accusative  See case.

active  An active sentence is one which has a basic pattern like the man is running or the dog bit the cat, i.e. it describes what one thing (the subject) does, often to another thing (the object). The verb in an active sentence can be said to be in the active voice. See also passive.

adjective  A word which qualifies or further describes a noun or noun phrase. Examples are colourless and green which qualify ideas in Colourless green ideas sleep furiously. Adjectives can also appear after verbs like be, e.g. The apples were green.

adjunct theta-role  See theta role.

adverb  A word which qualifies or further describes a verb, adjective or adverb. Examples are furiously which qualifies the verb sleep in Colourless green ideas sleep furiously, or intensely which qualifies stared in He stared at me intensely. Adverbs can also qualify adjectives, e.g. astonishingly qualifies the adjective vivid in an astonishingly vivid colour, or other adverbs, e.g. extremely qualifies the adverb slowly in the phrase extremely slowly. Many English adverbs are formed from an adjective plus the ending -ly. Words like very, which can only qualify adjectives or adverbs but not verbs, are sometimes called adverbs, but are perhaps best put in a separate category.

affix  An affix is a morpheme which is added to a root morpheme in the formation of a word. In its broadest sense, an affix can be a prefix, a suffix, or an infix. More narrowly, infixes are sometimes treated separately. See also morphology.

affricative  An affricative is a phone which can be thought of as a very rapid, blended sequence of a stop and a fricative. The stop and fricative must be produced in a very similar positions in the mouth. An English example is the 'ch sound' in choose, which is like a sequence of a 't sound' (a stop) and a 'sh sound' (a fricative). The phrases white shoes and why choose? sound very similar when spoken rapidly (but only in those dialects of English in which the [t] is not replaced by a glottal stop). In the IPA an affricative is represented by the corresponding stop symbol followed by the fricative symbol. It is important to note that the two symbols represent a SINGLE phone.

Agent  See theta role.

agreement  The syntax of a natural language often requires some words in a sentence to share certain grammatical features, which can show up as changes in the morphology of the words. This is called agreement; the words are said to agree in the relevant feature(s). For example, in English, determiners and nouns must agree in number within a noun phrase. Thus this cat is acceptable since this and cat are singular, but these cat is unacceptable since these is plural but cat is singular.

allophone  Each of the set of phones which correspond to a single phoneme of a language is called an allophone. Allophones of the same phoneme generally occur in different contexts and never distinguish one word from another. As an example, the 't sounds' in tea and tree constitute allophones of one English /t/ phoneme. The production of the two sounds differs in that speaker's tongue is in a slightly different place. A speech spectrograph will show a resulting sound difference. However, no English words differ ONLY in the substitution of one of these 't sounds' for the other. Allophones are written in square brackets (e.g. [t]) where it is necessary to distinguish them from phonemes (e.g. /t/).

alveolar  A phone produced when the tongue touches the tooth ridge behind the teeth (alveolus). See the diagram of a head for the location of the tooth ridge. The 't sound' in English is an alveolar stop, produced by stopping and then releasing the air flow out of the mouth by closing the tongue onto the tooth ridge.

anaphora  Some words in a sentence have little or no meaning of their own but instead refer to other words in the same or other sentences. This process is called anaphora. Pronouns are a good example. Consider the sentences: London had snow yesterday. It fell to a depth of a metre. To understand the second sentence it is necessary to identify it with snow rather than London or yesterday. English allows various forms of anaphora with verbs. For example, in I wanted to finish today, but I couldn't do it, the words do it refer to finish today and hence can be called anaphoric.

approximant  An approximant is a phone in which the tongue partly closes the airway, but not enough to cause a fricative. Examples in English are the phones that begin lap and woo. Approximants can be divided into liquids and glides. Approximants (especially glides) have some similarities to vowels.

argument theta-role  See theta role.

article  In English, a / an and the are called the indefinite and definite articles respectively. See also determiner.

aspect (of a verb)  Verbs can show not only the time location of an action (by grammatical tense), but also features such as whether the action is thought of as completed or continuing. A change in a verb which shows such a feature is often called an aspect of the verb. Compare ate with was eating in He ate rapidly when I came in and He was eating rapidly when I came in. Both refer to events in the past time; the difference lies in the implied relationship between the actions of 'eating' and 'coming in'. Syntactically, English has two marked aspects: progressive and perfect. The progressive aspect is formed by using the auxiliary be and the verb ending -ing. For example, I am eating it now implies both that the time is the present and that the 'eating' is currently in progress. The perfect aspect is formed by using the auxiliary have and the appropriate verb ending (usually -en or -ed): e.g. I have eaten it now, which implies both that the time is the present and that the 'eating' is finished. An English verb can show no aspect (e.g. runs or ran), progressive aspect (e.g. is running or was running), perfect aspect (e.g. has run or had run) or both perfect and progressive aspects (e.g. has been running or had been running). The table below shows the possible combinations of tense and aspect in English verbs.

Tense
PresentPast
AspectNoneI runI ran
ProgressiveI am runningI was running
PerfectI have runI had run
Perfect ProgressiveI have been runningI had been running

aspiration  If a phone is accompanied by a 'puff of air' it can be said to be aspirated. The 'p sound' in the English word pit is aspirated and is thus slightly different from the 'p sound' in spit, which is not aspirated.

assimilation  Particularly in rapid speech there is a tendency for neighbouring phones to become more similar, presumably to make pronunciation easier. For example, although the words Aston and Asda are both written with an s, the second word is normally pronounced as if spelt Azda. The reason seems to be that [s] and [t] are both voiceless, whereas [z] and [d] are both voiced. The sequence fricative followed by stop is easier to say if both have the same voicing.

ATN = Augmented Transition Network.

auxiliary  In English, one of a small set of verb-like words which can precede a main verb in a verb phrase. The auxiliaries and verbs are sometimes said to form a 'verb group' or 'compound verb'. Examples of auxiliaries are do in I really do not know, or may in I may see him tomorrow. Auxiliaries have verb-like properties, and may show changes in number, person and tense. Some words (e.g. have) can be either an auxiliary (e.g. I have seen him) or a verb (e.g. I have a car).

B

bilabial  A phone produced by the closure or partial closure of both lips. See the diagram of a head. The English sounds represented by the letters p in pit and b in bad are bilabial stops, produced by stopping and then releasing the air flow out of the mouth by closing the lips. Bilabial and labiodental phones are together classed as labial.

C

case  Nouns, noun phrases and pronouns play different syntactic roles in sentences. These roles correspond to changes of case in many languages. Consider, for example, the sentences She saw him and He saw her. The words she and he are used when they form the subject of the sentence and are said to be in the nominative case. She and he must be changed to her and him respectively when they form the object of the sentence and are said to be in the accusative case. Changes due to case are restricted to pronouns in English, but in other languages (e.g. Russian, Modern Greek), most nouns, pronouns, articles, adjectives, etc. will vary according to case.

circumstantial theta-role  See theta role.

consonant  (1) A phone which is produced other by allowing lung air to pass over the vibrating vocal cords and then freely out of the mouth, i.e. a phone other than a vowel. Consonants include stops, fricatives, affricatives and approximants. (2) A letter of the alphabet usually pronounced using a consonant phone is also called a consonant.

Be careful to distinguish these two usages. In a language with non-phonemic spelling, such as English, they can be quite different. The word mute, for example, begins with a single consonant letter, but in many British English dialects is pronounced with two opening consonant phones ([m] and [j] in IPA).

D

dental  A phone produced when the tongue touches the teeth. See the diagram of a head. The English sounds beginning the words this and think are alveolar fricatives, produced by partially stopping the air flow out of the mouth by touching the tongue on the teeth.

derivational morphology  See morphology.

determiner (det)  The definite and indefinite articles plus a small set of other similar words (e.g. genitive pronouns) which qualify nouns or noun phrases can be grouped as determiners. Examples of determiners are this, that, my. An English noun phrase always contains at most one determiner; singular noun phrases generally require exactly one determiner. Semantically, they determine that a particular instance of the noun is being referred (back) to. For example, There's a man at the door -- the word a introduces a man into the conversation. Tell the man I'll come in a minute -- the word the refers back to the previously mentioned man.

Noun phrases in the genitive act as determiners. Thus in I saw the old lady's cat, the genitive noun phrase the old lady's can be replaced by the single word determiner her.

dialect  Generally dialects of a language are more similar than different languages. However, what is a dialect and what is a language is often a political rather than a linguistic question. The division of Serbo-Croat, the common language of former Yugoslavia, into two languages, Serbian and Croatian, shows this rather sharply. A further example of very similar languages which might be called dialects of the same language are Dutch (spoken in the Netherlands) and Flemish (spoken in north-western Belgium). On the other hand, in China there are languages which are mutually un-intelligible when spoken but are often called dialects of one Chinese language. It is important to note that although some dialects have more social prestige in a country than others, this says nothing about their linguistic qualities.

diphthong  If the tongue moves significantly during the production of a vowel phone, the result is a diphthong. A diphthong sounds like a rapid, blended sequence of two separate vowels. An example in English is the vowel sound in the word kite, which is like a rapid combination of a kind of 'a sound' and a kind of 'i sound'. In the IPA a diphthong is represented by two vowel symbols. It is important to note that the two symbols represent a SINGLE phone.

direct object  See object.

E

ellipsis  A technical term for leaving out words in sentences. For example, in Brian ate the ice-cream and Judy the peaches, there is ellipsis, since the word ate is omitted after Judy.

F

feature  See semantic feature. (There are other uses of the term not covered here.)

feminine  See gender.

fricative  If during the production of a phone, air is made to pass through a narrow passage, a 'friction' sound or fricative is produced (i.e. a more-or-less 'hissing' sound). English examples are the 'f sound' in fee or the 'sh sound' in she.

G

gender  In some languages (but not English), nouns fall into a small number of classes which require changes in the articles, adjectives, etc. which qualify them. In Indo-European languages, these classes are traditionally called genders and labelled according to whether nouns for males (masculine gender), females (feminine gender) or neither (neuter gender) fall into these classes. French has two genders, masculine and feminine, shown for example by the use of le or la for the; German and Modern Greek have three genders, having neuter as well. Note that grammatical gender is not tied to biological sex, since, for example, the nouns meaning 'a young girl' are neuter in both German and Modern Greek. Thus as with number, grammatical gender is not the same as semantic gender.

genitive  See also case. Genitive is an alternative word for possessive, i.e. the genitive case marks the noun or pronoun as the possessor of something. In English, the genitive case of a noun is shown in writing by adding an s together with an appropriately positioned apostrophe. Thus of the boy becomes boy's, of the boys becomes boys'. The genitive or possessive pronouns are my, your, his, her, its [without an apostrophe!], our, their. Genitive noun phrases act as determiners.

glide  A glide is an approximant in which the tongue and lips move during the production of the sound. English examples are the initial phones in woo [w] and you [j].

glottal  A phone produced by closing or partially closing the vocal cords (or glottis). See the diagram of a head for the location of the vocal cords. The 'h sound' in English is a glottal fricative, produced by a strong air flow over partially open vocal cords.

grammar (1)  The word grammar is used as a collective word for morphology and syntax, i.e. for patterns both within and between words.

grammar (2)  The word grammar is also used a technical term for a rule-based approach which generates a particular set of sentences. Formally, a grammar consists of a set of nonterminal symbols (one of which is the start symbol), a set of terminal symbols and a set of productions or re-writing rules. Terminals (e.g. words) are the basic units of the sentences which the grammar generates. Nonterminals are symbols used only in the grammar itself. A production is a rule which says that the symbols on the left-hand side can be re-written as those on the right-hand side. One of the nonterminals must be the start symbol, i.e. the symbol from which re-writing starts.

grapheme  A grapheme is a 'spelling unit'. For example, in Spanish the combination ll represents a different sound from a single l. Thus these are two graphemes. In English, graphemes may be quite complex. For example -tion behaves more-or-less as a single grapheme in words like function.

H

I

idiolect  The language used by one individual is sometimes called an idiolect. A dialect or language can then be regarded as a collection of mutually intelligible idiolects.

indirect object  See object.

Indo-European  Linguists divide languages into a number of families, based on similarity and shared descent. Indo-European languages were natively spoken in a broad band through Europe to northern India and Bangladesh. Historically, the only major non-Indo-European languages spoken in this area were Finnish/Estonian, Hungarian, Basque and Turkish. It is believed that all the Indo-European languages are descended from one language spoken around 4,000 BC. It is important to be aware that different language families may be based on quite different principles, both in their sounds and in their grammar.

infix  A strong definition of an infix might be a morpheme which is added inside a root morpheme in the formation of a word. In a language like English, infixes, so defined, do not occur, since the root morpheme is indivisible. In Semitic languages, the root morpheme consists only of consonants -- usually three in, e.g., Arabic or Hebrew. A particular set of vowels and/or affixes combines with this root to form a word. Thus in Hebrew the root sgr has a basic meaning connected with "close" or "closed". Adding the vowels -a-a- (where the dashes mark the position of the root consonants) forms the active verb sagar; adding the prefix ni and the vowel --a- forms the passive verb nisgar. The inserted components can be called infixes, so that nisgar = prefix ni + infix --a- + root sgr.

A weaker definition of an infix might be one or more morphemes which are added inside a word to form another word. Such infixes are said to occur in English since, in colloquial speech, swear words can be inserted into other words, e.g. I hate this bloody university can become I hate this uni-bloody-versity. In English, such 'infixes' can apparently only be inserted before a stressed syllable.

See also morphology.

inflectional morphology  See inflection and morphology.

inflection  A grammatical change in the form of a word (more accurately of a lexeme), which leaves the 'base meaning' and the grammatical category of the word unchanged. In English, inflections are restricted to the endings of words (i.e. suffixes). Other languages may show changes elsewhere. As an example, the suffix s is the usual written plural inflection in English. Inflections in nouns may show changes of number, gender, case, etc.; in verbs, of number, person, tense, aspect, etc. See also morphology.

intonation  Intonation refers to changes in the tone or frequency of sounds during speech. For example, in English the tone usually falls at the end of a statement and rises at the end of a question, so that You want some coffee. and You want some coffee? can be distinguished by tone alone. In some languages (e.g. Chinese, Thai), sequences containing the same phones but with different intonation patterns correspond to different words.

IPA  The International Phonetic Alphabet or IPA is a set of symbols which can be used to represent the phones and phonemes of natural languages. A subset which can be used to represent 'Standard English English' (roughly the dialect of middle-class people from the south east of England) is given in a separate table.

J

K

L

labial  Bilabial and labiodental phones are together classed as labial.

labiodental  A phone produced by the partial closure of the lower lip on the upper teeth. See the diagram of a head. The English sounds represented by the letters f in fit and v in van are labiodental fricatives, produced by restricting the air flow out of the mouth by touching the lower lip on the upper teeth. Bilabial and labiodental phones are together classed as labial.

language  See natural language and dialect.

length  Length refers to the time duration of a phone. The English words beat and bead differ the length of the vowel as well as the voicing of the terminal stop; the vowel is longer in bead than in beat. In some languages the length of consonants may also be important.

lexeme  The four words eat, eats, eating and eaten are morphological variants of the word eat. The past tense ate is not so obviously morphologically connected to eat, but nevertheless has the same underlying meaning. Thus we may say that the five words eat, eats, eating, eaten and ate form a single lexeme, i.e. a single 'meaning entity'. A dictionary would be expected to contain only one definition for all five words. A lexeme is thus equivalent to what is often called a 'head word' in a dictionary.

lexicon  Often used as a technical term for the list of words and their types which is used with a grammar.

liquid  A liquid is a kind of approximant. English examples are the initial phones in lap and rap.

Location  See theta role.

M

masculine  See gender.

mood  A verb may be in one of several moods. The 'base' mood of a verb is the indicative or declarative, where the verb (and hence the sentence which contains it) states what is the case. The imperative mood is used to give instructions or commands. Compare The cat chases the mouse (indicative) with Chase the mouse! (imperative). The subjunctive mood, used to show hypothetical conditions, is rarely shown grammatically in Modern English. (If I were to tell you rather than If I was to tell you is one of the few uses which are at all common.) Languages vary widely in their use of moods.

As with other properties of verbs, it is important to distinguish between grammatical form and meaning. In the sentence I had finished my coursework before John came home, had finished is indicative in meaning, showing that the action was completed in the past before John came home. In the sentence If I had finished my coursework, I would have got a better mark, had finished is subjunctive in meaning, showing that the action was never completed.

morphology  The structure of words and the study of this structure. For example, a morphological analysis of the English word unknowingly might yield four components, called morphemes. These are the root know and three affixes, the prefix un indicating negation, and two suffixes ing and ly. Note that both spelling and pronunciation changes can take place when morphemes are combined. Thus the root happy plus the affix ly yields happily not *happyly. Many English words appear to contain morphemes, but resist neat division. For example, the suffix ish often indicates that the word refers to a language (e.g. English, Spanish, Danish, Swedish), but removing the suffix does not always leave a clear root morpheme (e.g. Spanish = ?Span(e) + ish). In other cases, it may be that a word was in the past created from distinct morphemes, but that this is not obvious to a contemporary speaker as the morphemes are no longer used in forming new words.

When an affix morpheme is an inflection, the word can be said to show inflectional morphology. Thus the word chased (= chase + ed) shows inflectional morphology. In many languages, including English, inflectional morphology is relatively predictable, and can be handled by rules.

In other cases, the word can be said to show derivational morphology. Thus the word output = out + put shows derivational morphology: adding the prefix out to the verb put creates a noun with the approximate meaning "that which was put out". In many languages, including English, derivational morphology is unpredictable, and so cannot easily be handled by rules. Thus there's no noun *outgo meaning "that which went out" (although there is a noun, most often used in the plural, outgoings = out + go + ing + s).

MT = Machine Translation

N

nasal  A nasal is a phone made by allowing air to flow out of the nose while possibly stopping it in the mouth. Allowing air to flow out of the mouth is achieved by opening the uvula (see the diagram of a head). English has three such phones: the nasal stops which end the words rum, run and rung.

In many languages (e.g. French, Punjabi), there are also nasal vowels, produced by allowing air to flow out of both the mouth and the nose.

natural language  Any language naturally used by people, i.e. not a man-made language like a programming language or Esperanto.

neuter  See gender.

NL = Natural Language.

NLP = Natural Language Processing.

nominative  See case.

nonterminal  See grammar.

noun  Semantically, a noun can be described as a word standing for the 'name of something.' A more useful test is that a noun or a noun phrase can be replaced by a pronoun, e.g. it or her. Examples of nouns are people, cats and intelligence in Many people think that cats have considerable intelligence. The strings of words many people and considerable intelligence are noun phrases in this example.

NP = Noun Phrase. See also phrase.

number  In English, nouns and verbs can be described as singular or plural, generally depending on whether the reference is to one or to many. Thus in the cat runs, cat is singular as is runs, whereas in cats run, cats is plural as is run. English nouns are generally clearly marked as singular or plural; verbs are clearly singular only in the third person singular of the present tense. However, grammatical number must be distinguished from semantic number; trousers is grammatically plural in English (since e.g. we must say my trousers are here and not *my trousers is here), but is clearly semantically singular. Some languages have dual number as well as singular and plural. For example, in Arabic, a special form of the noun corresponds to two rather than one or many. Other languages lack grammatical number (e.g. the Chinese languages).

O

object (of a sentence)  The direct object of an active sentence is a noun, noun phrase or pronoun which suffers the action of the verb. Thus in Those people dislike cats, cats is the object of the sentence. In English, only pronouns show case, and become accusative when forming the object of a sentence: thus, e.g., cats in the sentence above must be replaced by them rather than they. In other languages, nouns, adjectives, articles, etc. may all change case. The indirect object of a sentence in English is a noun or equivalent which, if the sentence were re-worded, would require a to (or sometimes a for). Thus in Your mother gave my brother a cake, a cake is the direct object and my brother the indirect object, since if we reverse brother and cake we need a to giving Your mother gave a cake to my brother. Direct and indirect objects may take different cases in some languages; e.g. in German, me is mich (accusative) when it is the direct object, but mir (dative) when it is the indirect object. See also subject.

P

palatal  A phone produced when the top of the tongue touches the hard palate. See the diagram of a head for the location of the hard palate. The English sounds represented by the letters sh in ship and s in measure are palatal fricatives, produced by partially stopping the air flow out of the mouth by touching the top of the tongue on the hard palate.

parse  To analyse a sentence using a grammar, including deciding whether it is valid and what its structure is according to the grammar.

participant theta-role  See theta role.

passive  A passive sentence is one which has a basic pattern like The cat was killed or The cat was killed by the dog, i.e. it describes what one thing (the subject) has done to it, often by another thing. The verb in an passive sentence can be said to be in the passive voice. See also active.

Patient  See theta role.

person (of a verb)  Verbs (in Indo-European languages at least) often vary depending on whether the subject of the verb is in the first person (singular = I, plural = we), the second person (singular and plural = you in modern English), or the third person (singular = he, she or it, plural = they). Only the verb be in the singular shows a full set of changes due to person in modern English: I am, you are, it is.

phone  A phone is a 'unit sound' of a language in the sense that it is the minimal sound by which two words can differ. For example, the English word feed contains three phones since each can be independently substituted to form a different word. In the IPA, the three phones can be written as [f], [i] and [d]. Examples of substitutions are: [fid] - [f] + [s] gives [sid], i.e. seed; [fid] - [i] + [u] gives [fud], i.e. food; [fid] - [d] + [t] gives [fit], i.e. feet. The whole of each phone must be substituted to change one word into another. It is important to note that whether or not speakers can distinguish between sounds is not a test of whether they constitute distinct phones. The word tea could be represented as [ti] and the word tree as [tri]. However, the two 't sounds' are not quite the same: the tongue is further back in the mouth when pronouncing the [t] in [tri] than when pronouncing the [t] in [ti]. How far to divide up sounds into phones is essentially a pragmatic question. Using more phones will enable speech to represented more accurately but at a cost in terms of complexity. See also allophone, phoneme.

phoneme  A phoneme is a minimally distinctive set of sounds in a language; sound sequences which differ in a single phoneme can constitute different words. Thus the pairs tip-dip and trip-drip show that English has two distinct phonemes, which we can write as /t/ and /d/, since substituting one for the other produces a different word. However, the pronunciation of /t/ (and /d/) is not the same in each pair: the tongue is further back in the mouth when /t/ is followed by /r/. Hence there are at least two phones corresponding to the /t/ phoneme. However there are no two English words in which the ONLY difference is that the 't sound in trip' is replaced by the 't sound in tip' -- these two sounds are allophones of the same phoneme. English speakers do not need to recognize the difference between them.

phonetics  Phonetics is the study of the sounds of speech (i.e. the study of phones). It can be distinguished from phonology which is more concerned with the underlying theory (i.e. the phonemes which underlie phones and the rules which govern the conversion of phonemes to phones and vice versa).

phonological rule  At some theoretical level, words can be considered to be composed of phonemes. The actual sound of a word then depends on which allophone is chosen for each phoneme. The context-sensitive rules which determine this are called phonological rules. Thus the word input can be considered to contain the phoneme /n/. However in fast speech in many dialects of English, the phone used will be [m]. The relevant phonological rule for English is that a nasal becomes articulated at the same position as a following stop.

phonology  See phonetics.

phrase  A string of words can often act as an exact grammatical substitute for a single word; such a string is called a 'phrase'. Thus e.g. a noun can be replaced by a noun phrase -- compare Whiskers is over there with That appalling pet of yours is over there, in which That appalling pet of yours is a noun phrase equivalent to the noun Whiskers.

plosive  See stop.

plural  See number.

possessive  See genitive.

pragmatics  A technical term meaning, roughly, what the person speaking or writing actually meant, rather than what the words themselves mean.

prefix  A prefix is a morpheme which is added before a root morpheme in the formation of a word. See morphology.

preposition  A preposition is one of a finite set of words (e.g. at, from, by) which in English must usually be followed by a noun or its equivalent. A prepositional phrase (PP) consists of a preposition followed by a noun, pronoun or noun phrase. Two major uses of prepositional phrases are to show location (e.g. on the mat in the cat sits on the mat) and motion (e.g. into the house in the cat runs into the house). The word preposition comes from pre plus position. In other languages (e.g. Japanese), there are postpositions: words which come after a noun or its equivalent.

production  See grammar.

pronoun  A pronoun is one of a small set of words which can substitute for a noun or noun phrase. It usually refers back to a previous occurrence of the noun or noun phrase. Thus, e.g., it in the previous sentence is a pronoun which refers back to A pronoun in the sentence before. The process of referring is sometimes called anaphora.

Q

R

Recipient  See theta role.

referential semantics  A system where the meaning of a word just is the thing it refers to.

RTN = Recursive Transition Network.

S

semantic feature  A semantic feature is a 'primitive' which a language processor (human or computer) is assumed to be able to determine independently of the language system. The meaning of words such as nouns or adjectives can then be described in terms of sets of these features. For example we might describe the meaning of words such as boy, man, girl and woman in terms of the features YOUNG, MALE and HUMAN. Boy would be [+YOUNG, +MALE, +HUMAN], woman would be [-YOUNG, -MALE, +HUMAN].

semantics  Used as a technical term for the meaning of words and sentences (see also pragmatics).

singular  See number.

start symbol  See grammar.

stop  Some phones are produced by completely stopping and then releasing the flow of air out of the mouth. These sounds are called stops. In most dialects of English there are three stop positions, corresponding to the initial phones in pale, tale and kale, or the terminal nasal phones in rum, run and rung. Some dialects of English (for example those spoken in England around London) also have a glottal stop, used, for example, instead of the 't sound' in words like bottle.

[The current tendency is to use the term plosive instead of stop. I resist this for the following reason. A stop actually consists of two phases: closure (when air pressure builds up) and release (when air explodes out). These phases can be separated: in many languages, stops are often not released (e.g. in the Bahasa language spoken in Malaysia and Indonesia, or at the end of words in many dialects of English spoken in northern England). The term 'unreleased stop' makes sense, whereas an 'unreleased plosive' is a contradiction.]

stress  Words can be divided into syllables, usually centred around a vowel. In many languages, including English, the duration and relative loudness of a syllable -- its stress -- are important. Thus only stress distinguishes the noun PROcess (as in the sentence This process is called assimilation) from the much less common verb proCESS (as in the sentence I usually process at the degree ceremony). The noun is stressed on the first syllable, the verb on the second.

STT = Speech To Text.

subject (of a sentence)  The subject of a sentence is the noun or noun equivalent which governs the verb, in the sense that if the language has agreements, the verb has to agree with the subject in number (as in English) or in gender (as in Arabic). Thus in English we have to say The dog chases the cats not The dog chase the cats; the verb chases agrees in number with the subject the dog rather than the object the cats. In the semantically equivalent passive sentence, The cats are chased by the dog, the fact that the cats is now the subject is shown by the need to use the plural auxiliary, are.

In an active sentence, the subject is often the entity which performs the action of the verb; in a passive sentence the subject is the entity which is in some sense the recipient of the action.

See also object.

suffix  A suffix is a morpheme which is added after a root morpheme in the formation of a word. See morphology.

syntax  The syntax of a language comprises, roughly speaking, the patterns into which its words can be validly arranged to form sentences. The combination of morphology and syntax is sometimes called the grammar of a language.

T

tense (of a verb)  The tense of a verb specifies the time at which its action occurs. The clearest examples in English are the present and past tenses. When saying I am eating an apple the speaker refers to the present; when saying I was eating an apple, s/he refers to the past. In its morphology, an English verb shows tense and aspect independently (see the table under aspect). Semantically, tense and aspect are not so easy to separate in English. I have eaten the apple is described morphologically as 'present perfect', but semantically is partly a reference to the past, and partly a reference to the action's being complete rather than continuing.

The future tense in English is formed by the use of the auxiliary will or sometimes shall. Morphologically, these auxiliaries can show 'past tense'; thus I would have been eating is the 'future past perfect progressive' of eat. Semantically, the combination of 'future' and 'past' is used to express 'conditionality', so that this form of the verb is usually called 'conditional'.

terminal node  A node in a transition network at which parsing can stop.

terminal  See grammar.

thematic role  See theta role.

theta role  (Often written as θ-role.) Verbs require a number of other components to be present in a sentence to complete their meaning. These components can be said to act as arguments to the verb, i.e. to be argument theta roles (or, alternatively, to play participant theta roles in relation to the action of the verb). For example, in the sentence The girl put the bottles on the table, the action of 'putting' involves three necessary thematic roles. These are Agent, the entity doing the putting; Patient, the entity which suffers the action of being put; and Location, where the Agent puts the Patient. A sentence containing the verb put will involve these three roles, even if they occur in different positions due to the syntax of the sentence. Thus exactly the same entities play exactly the same theta roles in the sentence The bottles were put on the table by the girl although the syntax is different from the previous sentence.

Another common θ-role is Recipient, the entity which receives something, typically the Patient. Thus the boy is the Recipient in both The girl gave the bottles to the boy and The boy was given the bottles.

In addition to argument or participant theta roles, there are adjunct or circumstantial theta roles. These show additional, non-required components. For example, in the kitchen plays an argument theta role in He was putting apples in the kitchen but only an adjunct theta role in He was eating apples in the kitchen. In both cases in the kitchen is a location, but put requires this role, eat merely allows it to be present.

TN = Transition Network.

TTS = Text To Speech.

U

unvoiced  See voicing.

V

velar  A phone produced when the top of the tongue touches the soft palate or velum. See the diagram of a head for the location of the soft palate. The English sounds represented by the letters k in kit and g in got are velar stops, produced by stopping and then releasing the air flow out of the mouth by touching the top of the tongue on the soft palate.

verb  A verb is traditionally described as a 'doing' word; thus in the sentences Colourless ideas sleep furiously and The dog bit the cat, sleep and bit are verbs. A more useful test is that a verb combines with an auxiliary in structures such as I can _ or I can _ them. English makes extensive use of 'verb groups' or 'compound verbs', such as has been eating in He has been eating fish in which one or more auxiliaries is combined with a verb.

Verbs may show a wide range of grammatical properties, including gender, person, tense, aspect, voice and mood. There are major differences among languages in the way these properties are shown grammatically and in their associated meanings.

voice  A verb may be in the active or passive voice, and hence so may the sentence in which the verb appears. Compare The dog chased the cat (active) with The cat was chased by the dog (passive). This use of the term 'voice' has no connection with 'voiced' or 'voiceless'.

In English, the grammatical voice of a verb is closely related to its meaning. In a sentence with an active verb the subject is typically the Agent, whereas in a sentence with a passive verb the subject is typically the Patient or the Recipient. Compare the active sentence The girl gave her mother a present, in which the girl is the subject and the Agent, with the passive sentences A present was given to her mother, in which a present is the subject and the Patient, and Her mother was given a present, in which her mother is the subject and Recipient.

In other languages, grammatical voice and meaning are less well aligned. For example, in Greek, both Classical and Modern, verbs which are passive in form may have active meanings, usually when the agent and patient are the same. Thus χτενίζω means "I comb" or "I am combing" but is only used when the patient (the thing being combed) is not the subject. The passive form χτενίζομαι normally means "I am combing [my hair]" rather than "I am being combed".

voiced  See voicing.

voiceless  See voicing.

voicing  Voicing refers to whether or not the vocal cords are vibrated during the production of a phone. Phones such as vowels or [b] or [d] in which the vocal cords are vibrated are said to be voiced. Phones such as [s] or [p] in which the vocal cords are not vibrated are said to be voiceless or unvoiced.

vowel  (1) A phone which is produced by allowing lung air to pass over the vibrating vocal cords and then freely out of the mouth. Thus vowels can be continued until you run out of breath. The positions of the lips and tongue alter the size and shape of the resonating cavity to produce different sounds. (2) A letter of the alphabet usually pronounced using a vowel phone is also called a vowel.

Be careful to distinguish these two usages. In a language with non-phonemic spelling, such as English, they can be quite different. The word site, for example, contains two vowel letters but only one vowel phone since the terminal e is not pronounced.

See also consonant.

W

X

Y

Z

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z