Inflection

Regular and irregular inflection

When a given word class is subject to inflection in a particular language, there are generally one or more standard patterns of inflection (the paradigms described below) that words in that class may follow. Words which follow such a standard pattern are said to be regular; those that inflect differently are called irregular.

For instance, many languages that feature verb inflection have both regular verbs and irregular verbs. In English, regular verbs form their past tense and past participle with the ending -[e]d. Therefore, verbs like play, arrive, and enter are regular, while verbs like sing, keep, and go are irregular. Irregular verbs often preserve patterns that were regular in past forms of the language, but which have now become anomalous; in rare cases, there are regular verbs that were irregular in past forms of the language.

Other types of irregular inflected form include irregular plural nouns, such as the English plurals mice, children, and women and the French yeux (the plural of œil, 'eye'); and irregular comparative and superlative forms of adjectives or adverbs, such as the English better and best (which correspond to the positive forms good or well).

Irregularities can have four basic causes:

1. Euphony: Regular inflection would result in forms that sound esthetically unpleasing or are difficult to pronounce (English far → farther or further; Spanish tener → tengo, tendré vs. comer → como, comeré; Portuguese vs. Spanish andar (in both languages) → Portuguese andaram vs. Spanish anduvieron).
2. Principal parts: These are generally considered to have been formed independently of one another, so they must be memorized when learning a new word and its inflections. Example: Latin dīcō, dīcere, dīxī, dictum → Spanish digo, decir, dije, dicho.
3. Strong vs. weak inflection: In some cases, two inflection systems exist, conventionally classified as "strong" and "weak". For instance, English and German have weak verbs that form the past tense and past participle by adding an ending (English jump → jumped; German machen → machte); versus strong verbs that change vowel, and in some cases form the past participle by adding -en (English swim → swam, swum; German schwimmen → schwamm, geschwommen). Ancient Greek verbs are likewise said to have had a first aorist (ἔλῡσα) and a second aorist (ἔλιπον).
4. Suppletion: The "irregular" form was originally derived from a different root (English person → people). The comparative and superlative forms of good in many languages display this phenomenon (e.g. English good, better, best).

Declension and conjugation

Two traditional grammatical terms refer to inflections of specific word classes:

• Inflecting a noun, pronoun, adjective, adverb, article, or determiner is known as declining it. The forms may express number, case, gender, or degree of comparison.
• Inflecting a verb is called conjugating it. The forms may express tense, mood, voice, aspect, person, or number.

An organized list of the inflected forms of a given lexeme or root word is called its declension if it is a noun, or its conjugation if it is a verb.

Below is the declension of the English pronoun I, which is inflected for case and number.

The pronoun who is also inflected according to case. Its declension is defective, in the sense that it lacks a reflexive form.

The following table shows the conjugation of the verb to arrive in the indicative mood; suffixes inflect it for person, number, and tense:

The non-finite forms arrive (bare infinitive), arrived (past participle), and arriving (present participle, gerund), although not inflected for person or number, can also be regarded as part of the conjugation of the verb to arrive. Compound verb forms, such as I have arrived, I had arrived, or I will arrive, can be included also in the conjugation of the verb for didactic purposes, but they are not overt inflections of arrive. The formula for deriving the covert form, in which the relevant inflections do not occur in the main verb, is:

pronoun + conjugated auxiliary verb + non-finite form of main verb.

Conjugations of verbs in other languages are often much more complex, as illustrated below.

Inflectional paradigm

An inflectional paradigm refers to a pattern (usually a set of inflectional endings), in which a class of words follows the same pattern. Nominal inflectional paradigms are called declensions, and verbal inflectional paradigms are termed conjugations. For instance, there are five types of Latin declension. Words that belong to the first declension usually end in -a and are usually feminine, these words share a common inflectional framework. Meanwhile words that belong to the second declension end in -us, -um, or -er, and are usually masculine or neuter. In Old English, nouns are divided into two major categories of declension, the strong and weak ones, as shown below:

The terms strong declension and weak declension are primarily relevant to well-known dependent-marking languages (such as the Indo-European languages, or Japanese). In dependent-marking languages, nouns in adpositional (prepositional or postpositional) phrases can carry inflectional morphemes.

In head-marking languages, the adpositions can carry the inflection in adpositional phrases. This means that these languages will have inflected adpositions. In Western Apache (San Carlos dialect), the postposition -ká’ 'on' is inflected for person and number with prefixes:

Compared to derivation

Inflection is the process of adding inflectional morphemes that modify a verb's tense, mood, aspect, voice, person, or number or a noun's case, gender, or number, rarely affecting the word's meaning or class. Examples of applying inflectional morphemes to words are adding -s to the root dog to form dogs and adding -ed to wait to form waited.

In contrast, derivation is the process of adding derivational morphemes, which create a new word from existing words and change the semantic meaning or the part of speech of the affected word, such as by changing a noun to a verb.

Distinctions between verbal moods are mainly indicated by derivational morphemes.

Words are rarely listed in dictionaries on the basis of their inflectional morphemes (in which case they would be lexical items). However, they often are listed on the basis of their derivational morphemes. For instance, English dictionaries list readable and readability, words with derivational suffixes, along with their root read. However, traditional English dictionaries do not list books as an entry separately from book; the same goes for jumped and jump.

Inflectional morphology

Languages that add inflectional morphemes to words are called inflectional languages or inflected languages. Morphemes may be added in several different ways:

• Affixation, or simply adding morphemes onto the word without changing the root;
• Reduplication, repeating all or part of a word to change its meaning;
• Alternation, exchanging one sound for another in the root (usually vowel sounds, as in the ablaut process found in Germanic strong verbs and the umlaut often found in nouns, among others);
• Suprasegmental variations, such as of stress, pitch or tone, where no sounds are added or changed but the intonation and relative strength of each sound is altered regularly.

Inflection through reduplication

Reduplication is a morphological process where a constituent is repeated. The direct repetition of a word or root is called total reduplication (or full reduplication). The repetition of a segment is referred to as partial reduplication. Reduplication can serve both derivational and inflectional functions. A few examples are given below:

Inflection through tone change

An example is Tlatepuzco Chinantec (an Oto-Manguean language spoken in Southern Mexico), where tones are able to distinguish mood, person, and number:

Case can be distinguished with tone as well, as in Maasai language (a Nilo-Saharan language spoken in Kenya and Tanzania):

In various languages

Indo-European languages (fusional)

Because the Proto-Indo-European language was highly inflected, all of its descendant Indo-European languages are inflected to a greater or lesser extent. In general, older Indo-European languages such as Latin, Ancient Greek, Old English, Old Norse, Old Church Slavonic, and Sanskrit are extensively inflected because of their temporal proximity to Proto-Indo-European. Deflexion has caused modern versions of some Indo-European languages that were previously highly inflected to be much less so; an example is Modern English, as compared to Old English. In general, languages where deflexion occurs replace inflectional complexity with more rigorous word order, which provides the lost inflectional details. Most Slavic languages and some Indo-Aryan languages are an exception to the general Indo-European deflexion trend, continuing to be highly inflected (in some cases acquiring additional inflectional complexity and grammatical genders, as in Czech and Marathi).

English

Old English was a moderately inflected language, using an extensive case system similar to that of modern Icelandic, Faroese, and German. Middle and Modern English lost progressively more of the Old English inflectional system. Modern English is considered a weakly inflected language, since its nouns have only vestiges of inflection (plurals, the pronouns), and its regular verbs have only four forms: an inflected form for the past indicative and subjunctive (looked), an inflected form for the third-person-singular present indicative (looks), an inflected form for the present participle (looking), and an uninflected form for everything else (look). While the English possessive indicator 's (as in "Jennifer's book") is a remnant of the Old English genitive case suffix, it is now considered by syntacticians not to be a suffix but a clitic, although some linguists argue that it has properties of both.

Scandinavian languages

Old Norse was inflected, but modern Swedish, Norwegian, and Danish have lost much of their inflection. Grammatical case has largely died out with the exception of pronouns, just as in English. However, adjectives, nouns, determiners, and articles still have different forms according to grammatical number and grammatical gender. Danish and Swedish only inflect for two different genders while Norwegian has to some degree retained the feminine forms and inflects for three grammatical genders like Icelandic. However, in comparison to Icelandic, there are considerably fewer feminine forms left in the language.

By contrast, Icelandic preserves almost all of the inflections of Old Norse and remains heavily inflected. It retains all the grammatical cases from Old Norse and is inflected for number and three different grammatical genders. The dual-number forms from Old Norse are, however, almost completely lost.

Unlike in other Germanic languages, nouns are inflected for definiteness in all Scandinavian languages, as in the following case for Nynorsk Norwegian:

Adjectives and participles are also inflected for definiteness in all Scandinavian languages, as they were in Proto-Germanic.

Other Germanic languages

Modern German remains moderately inflected, retaining four noun cases, although the genitive started falling into disuse in all but formal writing in Early New High German. The case system of Dutch, simpler than that of German, is also further simplified in common usage. Afrikaans, recognized as a distinct language in its own right rather than a Dutch dialect only in the early 20th century, has lost almost all inflection.

Latin and the Romance languages

The Romance languages, such as Spanish, Italian, French, Portuguese, and especially – with its many cases – Romanian, have more overt inflection than English, especially in verb conjugation. Adjectives, nouns, and articles are considerably less inflected than verbs, but they still have different forms according to number and grammatical gender.

Latin, the mother tongue of the Romance languages, was highly inflected; nouns and adjectives had different forms according to seven grammatical cases (including five major ones) with five major patterns of declension, and three genders instead of the two found in descendant Romance tongues. Latin inflection included: four patterns of conjugation, in six tenses; three moods (indicative, subjunctive, imperative), plus the infinitive, participle, gerund, gerundive, and supine form; and two voices (passive and active). All were overtly expressed by affixes (passive voice forms were periphrastic, in three tenses).

Baltic languages

The Baltic languages are highly inflected. Nouns and adjectives are declined in up to seven overt cases. Additional cases are defined in various covert ways. For example, an inessive case, an illative case, an adessive case, and allative case are borrowed from Finnic. Latvian has only one overt locative case but it syncretizes the above four cases to the locative marking them by differences in the use of prepositions. Lithuanian breaks them out of the genitive case, accusative case and locative case by using different postpositions.

Dual form is obsolete in standard Latvian and nowadays it is also considered nearly obsolete in standard Lithuanian. For instance, in standard Lithuanian it is normal to say "dvi varnos (plural) – two crows" instead of "dvi varni (dual)". Adjectives, pronouns, and numerals are declined for number, gender, and case to agree with the noun they modify or for which they substitute. Baltic verbs are inflected for tense, mood, aspect, and voice. They agree with the subject in person and number (not in all forms in modern Latvian).

Slavic languages

All Slavic languages make use of a high degree of inflection, typically having six or seven cases and three genders for nouns and adjectives. However, the overt case system has disappeared almost completely in modern Bulgarian and Macedonian. Most verb tenses and moods are also formed by inflection (however, some are periphrastic, typically the future and conditional). Inflection is also present in adjective comparation and word derivation.

Declensional endings depend on case (nominative, genitive, dative, accusative, locative, instrumental, or vocative), number (singular, plural, or sometimes dual), gender (masculine, feminine, or neuter) and animacy (animate or inanimate). Declension in most Slavic languages also depends on whether the word is a noun or an adjective; this is a distinction unusual in other language families. Slovene and Sorbian languages use the uncommon dual number between singular and plural. For some words, dual survived also in Polish and some other Slavic languages. Modern Russian, Serbian, and Czech also use a more complex sub-form of plural as a separate inflectional pattern; often called "dual", this misnomer actually applies to numbers 2, 3, and 4, as well as larger numbers ending in 2, 3, and 4 (with the exception of the teens, which are handled as plural); thus, 2, 22, and 102 are "dual", but 7, 12 and 127 are plural).

In addition, in some Slavic languages, such as Polish, word stems are frequently modified by the addition or absence of endings, resulting in consonant and vowel alternation (apophony).

Arabic (fusional)

Modern Standard Arabic (also called Literary Arabic) is an inflected language. It uses a system of independent and suffix pronouns classified by person and number, with verbal inflections that also mark person and number. Suffix pronouns are used as markers of possession and as objects of verbs and prepositions. The kashida or tatweel (ـ) marks where the verb stem, verb form, noun, or preposition is placed.

Regional Arabic dialects (e.g. Moroccan Arabic, Egyptian Arabic, Gulf Arabic), used for everyday communication, tend to have less inflection than the more formal Literary Arabic. For example, in Jordanian Arabic, the second- and third-person feminine plurals (أنتنّ antunna and هنّ hunna) and their respective unique conjugations are lost and replaced by the masculine (أنتم antum and هم hum), whereas in Lebanese and Syrian Arabic, هم hum is replaced by هنّ hunna.

In addition, the system known as ʾIʿrāb places vowel suffixes on each verb, noun, adjective, and adverb, according to its function within a sentence and its relation to surrounding words.

Uralic languages (agglutinative)

The Uralic languages are agglutinative, following from the agglutination in Proto-Uralic. The largest languages in the family are Hungarian, Finnish, and Estonian – all European Union official languages. Uralic inflection is, or is developed from, affixing. Grammatical markers directly added to the word perform the same function as prepositions in English. Almost all words are inflected according to their roles in the sentence: verbs, nouns, pronouns, numerals, adjectives, and some particles.

Hungarian and Finnish, in particular, often simply concatenate suffixes. For example, Finnish talossanikinko, 'in my house, too?', consists of talo-ssa-ni-kin-ko. However, in the Finnic languages (Finnish, Estonian, etc.) and the Sami languages, there are processes which affect the root, particularly consonant gradation. The original suffixes may disappear (and appear only by liaison), leaving behind the modification of the root. This process is extensively developed in Estonian and Sami, and makes them also inflected as well as agglutinating languages. The Estonian illative case, for example, is expressed by a modified root: maja → majja (from early form *maja-han).

Altaic languages (agglutinative)

Though the proposed Altaic languages superfamily is widely considered by linguists to be simply a sprachbund of unrelated languages, three language families united as Altaic by a small subset of linguists – Turkic, Mongolic, and Manchu-Tungus – are agglutinative. The largest languages in this grouping are Turkish, Azerbaijani, and Uzbek, all of which are Turkic. Altaic inflection is, or is developed from, affixing. Grammatical markers directly added to the word perform the same function as stand-alone prepositions and such in English. Almost all words are inflected according to their roles in the sentence: verbs, nouns, pronouns, numerals, adjectives, and some particles.

Basque (agglutinative nominal inflection / fusional verb inflection)

Basque, a language isolate, is a highly inflected language, heavily inflecting both nouns and verbs.

Noun-phrase morphology is agglutinative and consists of suffixes which simply attach to the end of a stem (and more generally, only once at the very end of the nominal syntagma). These suffixes are in many cases fused with the article (-a for singular and -ak for plural), which in general is required to close a noun phrase in Basque if no other determiner is present, and unlike an article in many languages, it can only partially be correlated with the concept of definiteness. Proper nouns do not take an article, and indefinite nouns without the article (called mugagabe in Basque grammar) are highly restricted syntactically. Basque is an ergative language, meaning that inflectionally the single argument (subject) of an intransitive verb is marked in the same way as the direct object of a transitive verb. This is called the absolutive case and in Basque, as in most ergative languages, it is realized with a zero morph; in other words, it receives no special inflection. The subject of a transitive verb receives a special case suffix, called the ergative case.

There is no case marking concord in Basque; case suffixes, including those fused with the article, are added only to the last word in a noun phrase. Plurality is not marked on the noun and is identified only in the article or other determiner, possibly fused with a case marker. The examples below are in the absolutive case with zero case marking, and include the article only:

The noun phrase is declined for 11 cases: Absolutive, ergative, dative, possessive-genitive, benefactive, comitative, instrumental, inessive, allative, ablative, and local-genitive. These are signaled by suffixes that vary according to the categories of Singular, Plural, Indefinite, and Proper Noun, and many vary depending on whether the stem ends in a consonant or vowel. The Singular and Plural categories are fused with the article, and these endings are used when the noun phrase is not closed by any other determiner. This gives a potential 88 different forms, but the Indefinite and Proper Noun categories are identical in all but the local cases (inessive, allative, ablative, local-genitive), and many other variations in the endings can be accounted for by phonological rules operating to avoid impermissible consonant clusters. Local case endings are not normally added to animate Proper Nouns. The precise meaning of the local cases can be further specified by additional suffixes added after the local case suffixes.

Verb forms are extremely complex, agreeing with the subject, direct object, and indirect object; and include forms that agree with a "dative of interest" for intransitive verbs as well as allocutive forms where the verb form is altered if one is speaking to a close acquaintance. These allocutive forms also have different forms depending on whether the addressee is male or female. This is the only area in Basque grammar where gender plays any role at all. Subordination could also plausibly be considered an inflectional category of the Basque verb since subordination is signaled by prefixes and suffixes on the conjugated verb, further multiplying the number of potential forms.

Transitivity is a thoroughgoing division of Basque verbs, and it is necessary to know the transitivity of a particular verb in order to conjugate it successfully. In the spoken language only a handful of commonly used verbs are fully conjugated in the present and simple past, most verbs being conjugated by means of an auxiliary which differs according to transitivity. The literary language includes a few more such verbs, but the number is still very small. Even these few verbs require an auxiliary to conjugate other tenses besides the present and simple past.

The most common intransitive auxiliary is izan, which is also the verb for 'to be'. The most common transitive auxiliary is ukan, which is also the verb for 'to have'. (Other auxiliaries can be used in some of the tenses and may vary by dialect.) The compound tenses use an invariable form of the main verb (which appears in different forms according to the "tense group") and a conjugated form of the auxiliary. Pronouns are normally omitted if recoverable from the verb form. A couple of examples will suffice to demonstrate the complexity of the Basque verb:

Liburu-ak

Book-PL.the

saldu

sell

dizkiegu.

AUX.3PL/ABS.3PL/DAT.1PL/ERG

'We sold the books to them.'

Kafe-a

Coffee-the

gusta-tzen

please-HAB

zaidak.

AUX.ALLOC/M.3SG/ABS.1SG/DAT

'I like coffee.' More literally, 'Coffee pleases me.' (Used when speaking to a male friend.)

The morphs that represent the various tense/person/case/mood categories of Basque verbs, especially in the auxiliaries, are so highly fused that segmenting them into individual meaningful units is nearly impossible, if not pointless. Considering the multitude of forms that a particular Basque verb can take, it seems unlikely that an individual speaker would have an opportunity to utter them all in his or her lifetime.

Mainland Southeast Asian languages (isolating)

Most languages in the Mainland Southeast Asia linguistic area (such as the varieties of Chinese, Vietnamese, and Thai) are not overtly inflected, or show very little overt inflection, and are therefore considered analytic languages (also known as isolating languages).

Chinese

Standard (Mandarin) Chinese does not possess overt inflectional morphology. While some languages indicate grammatical relations with inflectional morphemes, Chinese utilizes word order and particles. Consider the following language-contrasting examples:

• Latin:
  • Puer puellam videt.
  • Puellam puer videt.

Both sentences mean 'The boy sees the girl.' This is because puer (boy) is singular nominative, puellam (girl) is singular accusative. Since the roles of puer and puellam have been marked with case endings, the change in position does not matter.

• Modern Standard Chinese:
  • 我给了他一本书; wǒ gěile tā yī běn shū; 'I gave him a book'
  • 他给了我一本书; tā gěile wǒ yī běn shū; 'He gave me a book'

The situation is very different in Chinese. Since Modern Chinese makes no use of inflection, the relationship between wǒ ('I' or 'me') and tā ('he' or 'him') can only be determined by their position.

English, in contrast to both of the above, uses word order/position to establish relationships between words and steer overall sentence meaning, but also reinforces this approach with inflection, which may permit limited word order changes.

In Classical Chinese, pronouns were overtly inflected to mark case. However, these overt case forms are no longer used; most of the alternative pronouns are considered archaic in modern Mandarin Chinese. Classically, 我; wǒ was used solely as the first-person accusative, while 吾; wú was generally used as the first-person nominative.

Certain varieties of Chinese are known to express meaning by means of tone change, although further investigations are required. Tone change is a morphologically conditioned alternation used as an inflectional or derivational strategy. It is distinguished from tone sandhi, a compulsory change that occurs when certain tones are juxtaposed. Examples of tone change from Taishanese and Zhongshan dialect (both Yue dialects spoken in Guangdong Province) are shown below (The superscripted numbers indicate the Chao tone numerals):

• Taishanese

• Zhongshan dialect

The following table compares the personal pronouns of Sixian dialect of Taiwanese Hakka with Zaiwa and Jingpho (both Tibeto-Burman languages spoken in Yunnan and Myanmar/Burma).

Japanese (agglutinative)

Japanese shows a high degree of overt inflection of verbs, less so of adjectives, and very little of nouns, but it is mostly strictly agglutinative and extremely regular. Fusion of morphemes may happen in colloquial speech, for example: the causative-passive 〜せられ〜 (-serare-) fuses into 〜され〜 (-sare-), as in 行かされる (ikasareru, 'is made to go'), and the non-past progressive 〜ている (-teiru) fuses into 〜てる (-teru) as in 食べてる (tabeteru, 'is eating'). Formally, every noun phrase must be marked for case, but this is done by invariable particles (clitic postpositions). (Many grammarians consider Japanese particles to be separate words, and therefore not an inflection, while others view them as agglutinative affixes, and therefore consider Japanese nouns to be inflected.)

Auxiliary languages

Some international auxiliary languages, such as Lingua Franca Nova, Glosa, and Frater, have no inflection. Other auxiliary languages, such as Esperanto, Ido, and Interlingua have comparatively simple inflectional systems. Some on the contrary, like Volapük, are highly inflected (though perfectly regular).

Esperanto

In Esperanto, an agglutinative language, nouns and adjectives are inflected for case (nominative or accusative) and number (singular or plural), according to a simple paradigm without irregularities. Verbs are not inflected for person or number, but they are inflected for tense (past, present, or future) and mood (indicative, infinitive, conditional, or jussive). They also form active and passive participles, which may be past, present, or future. All verbs are regular.

Ido

Ido has a different form for each verbal tense (past, present, future, volitive, and imperative) plus an infinitive, and both a present and past participle. There are, however, no verbal inflections for person or number, and all verbs are regular.

Nouns are marked for number (singular or plural), and the accusative case may be shown in certain situations, typically when the direct object of a sentence precedes its verb. On the other hand, adjectives are unmarked for gender, number, or case (unless they stand on their own, without a noun, in which case they take on the same desinences as the missing noun would have taken). The definite article la ('the') remains unaltered regardless of gender or case, and also of number, except when there is no other word to show plurality. Pronouns are identical in all cases, though exceptionally the accusative case may be marked, as for nouns.

Interlingua

Interlingua, in contrast with the Romance languages, has almost no irregular verb conjugations, and its verb forms are the same for all persons and numbers. It does, however, have compound verb tenses similar to those in the Romance, Germanic, and Slavic languages: ille ha vivite, 'he has lived'; illa habeva vivite, she had lived. Nouns are inflected by number, taking a plural -s, but rarely by gender (only when referring to a male or female being). Interlingua has no noun–adjective agreement by gender, number, or case. As a result, adjectives ordinarily have no inflections. They may take the plural form if they are being used in place of a noun: le povres, 'the poor'.