Notes on English inflection

In this post I discuss some features of English inflection:

• English identifies 6 combinations of person (eg 1^st person singular, 3^rd person plural) but mostly distinguishes only 2 categories: (a) 3^rd person singular versus (b) all others.
• the most prominent inflectional ending in English is -s, which serves as both:
  – the 3^rd person singular ending on verbs; and
  – the non-[3^rd person.singular] ending on nouns.
• phonetically, that inflectional ending behaves the same way as -’s in the ‘Saxon Genitive’ and also as the contracted form of the 3^rd person singular form of be (’s). That phonetic behaviour is also almost identical to the behaviour of the past tense verbal suffix -d.       
• in negative questions, the seemingly anomalous form aren’t I? instead of the expected amn’t I? As a result, the negative interrogative forms of be show yet another split between 3^rd person singular isn’t and all other forms aren’t

The rest of this post looks at:

• verbal inflection
• inflecting nouns
• the ’s possessive (Saxon genitive)
• past tense -d
• aren’t I
• do English verbs in fact inflect for number?
• one suffix or 3?

Verbal inflection

English verbs inflect for person, number and tense. Inflection for person and number occurs only in the present tense, not in the past tense.

Because there are 3 persons (1^st, 2^nd and 3^rd) and 2 numbers (singular and plural), it might seem that there should be 6 inflected forms in the present tense (3 singular and 3 plural). But in fact, for all verbs (except be) 5 of those 6 forms are identical. What’s more, all 5 of those forms have no suffix (or maybe a suffix that is zero).

So, in line with an analysis by Sag, Wasow and Bender (2003), it seems sensible to say that in the present tense English verbs have:

• one form for the 3^rd person singular. This form ends in the suffix -s (eg make–s).
• another form (with no suffix) for all other instances (eg make).

This statement could be written as a simple morphological rule for present tense verbs:

• 3.sing ↔ -s
• non-[3.sing] ↔ –

The only English verb with more than 2 forms in the present tense is be. It has 3 forms: am, is, are.

Inflecting nouns

It is striking how similarly verbs and nouns inflect for number:

• both categories (nouns and verbs) have 2 inflectional forms reflecting number. One form has the suffix -s and the other form (the default form) has no suffix.
• in each phrase containing a verb and its subject, the suffix -s appears exactly one time:
  – on the verb, if the subject is singular.
  – on the subject if the subject is plural. 

Table 1 shows these patterns

Table 1. Inflection of subjects and verbs

Thus, bearing in mind that nouns are typically only 3^rd person, we can unify the analysis of nouns and verbs. For nouns, we could set up morphological rules that are similar to those for verbs (leaving 1^st and 2^nd person pronouns aside for a minute):

• 3.sing ↔ –
• non-[3.sing] ↔ -s

In practice for nouns, non-[3.sing] would apply only to 3^rd person plural.

Table 2 summarises the suffixes for subject nouns (not pronouns) and verbs.

Table 2. Suffixes on subjects and verbs

Not all nouns form their plural with the suffix -s. Some plural nouns have other endings and some others take no ending. Nouns with no ending: include:

• plurals borrowed from other languages (eg criteria, cherubim)
• a few relics of old forms (eg mice, oxen, brethren)
• invariable words (eg sheep, fish).

Personal pronouns

The discussion so far covers nouns but not pronouns. For pronouns, the story is a bit more complex. In English, personal pronouns can mark more categories than nouns can:

• personal pronouns inflect for number (except in the 2^nd person) but also for person and case.
• 3^rd person singular pronouns also inflect for gender.
• nouns inflect only for number (and some nouns, eg sheep, do not even distinguish plural from singular).

Strikingly, a split between 3rd person singular (table 3) and all other forms (non-[3.SING] (table 4) exists here too, just as it does for present tense verbs.

Table 3. personal pronouns: 3^rd person singular

Table 4. Other personal pronouns

Thus, the only pronouns for which English morphological rules need to specify gender are 3^rd personal singular forms. The rules do not need to specify gender for any 1^st or 2^nd person personal pronouns, nor for any 3^rd person plural pronouns.

Gender
In this post, I analyse only the traditional English pronominal system that uses 3 sets of singular pronouns: male, female and neuter. This post does not analyse:
– how those pronouns are selected
– whether the number and content of those sets might change in the future
-whether any such changes would be desirable.  

Morphology of pronouns

I can see 2 ways to set up a morphological rule for pronouns:

• Approach A:
  step 1. Distinguish 3.sing from non-[3.sing]
  step 2. Split non-[3.sing] into 1^st, 2^nd and 3^rd
  step 3. Split 1^st person into singular and plural
• Approach B:
  step 1. Split into 1^st, 2^nd and 3^rd person.
  step 2. Split into singular and plural

Approach A would start with the same step as the single step needed for nouns, but it would require 3 steps.

Approach B would have only 2 steps but neither step would be the same as the single step needed for nouns.

The ’s possessive (Saxon genitive)

There is another English form that has the suffix -s, just like the 3^rd person singular suffix for verbs and the regular plural suffix for nouns. That suffix is the possessive -s (the ‘Saxon genitive’). Please see:

• for more on the Saxon Genitive https://languagemiscellany.com/2024/09/saxon-genitive/
• for an argument that the Saxon Genitive is perhaps best viewed as a clitic attaching at the end of a phrase, rather than as a suffix added to a noun Does English really have case? – Language Miscellany
• for a similar construction in Swedish https://languagemiscellany.com/2023/10/surprised-by-genitive-s-in-swedish/

All 3 suffixes exhibit similar phonetic behaviour: they surface as:

• [s] after unvoiced consonants;
• [z] after voiced consonants or after vowels; and
• [-iz] after sibilants.

After a final sibilant suffix attaching to names, the Saxon Genitive coalesces optionally with the sibilant instead of surfacing as [-iz] eg Burns or Burns’s.  (Berg 2011)

Contracted form of is

A 4^th item that shows similar phonetic behaviour is the contracted form ’s of the third singular copula is.

Past tense –d

Alongside the 3 suffixes (or clitics) in -s (or –’s) discussed above, one other English suffix is the past tense -d. There are some important similarities between the 3 versions of -s and past tense -d. Both [-s] and [-d]:

• are dental consonants, produced with the tip of the tongue touching the back of the top teeth, or the alveolar ridge just above the top teeth.
• have 3 allophones, appearing: –
  – after unvoiced consonants [-s] / [-t];
  – after voiced consonants [-z] / [-d];
  – after sibilants [-iz] / [-id]
• are ignored in applying some ‘phonotactic’ constraints that limit the possible clusters of consonants at the end of a syllable.

On the other hand, there are 2 differences between the -s (or –’s) suffixes and the -d suffix:

• [-s] and its allophones are fricatives (they let air flow throughout the production of the sound) but [-d] and its allophones are stops (they block the air flow completely before then releasing it).
• although [-s] and [-d] behave similarly and both have an unvoiced allophone and 2 voiced allophones, [-s] is always written with a letter for an unvoiced consonant; on the other hand, [-d] is always written with a letter for a voiced consonant.

Aren’t I?

I’ve already mentioned 2 places where English distinguishes 3^rd person singular from all 5 other combinations of person (1^st, 2^nd, 3^rd) and number (singular, plural):

• in the verbal suffixes for the present tense; and
• in the gender system for personal pronouns.

A similar distinction occurs in a 3^rd place: the inverted negative verbal forms of be used for negative questions (table 5): isn’t is used for 3^rd person singular and aren’t is used for all 5 other forms. (‘inverted’ means that the verb proceeds the subject, instead of following it as it does in affirmative sentences.)

In contrast, 3 forms are used in the affirmative: is for 3^rd person singular, am for 1^st person singular and are for all 4 other forms.

Table 5. Present tense of be

Perhaps it is just coincidence that the same split between 3^rd person singular and all 5 other combinations of person number appears here too.

Also, there is another way of looking at the seemingly anomalous form aren’t I?. Perhaps it is related to a similar pronunciation pattern in the modal verbs can and shall, and even in will (table 6).

Table 6. Affirmative and negative questions: be and 3 modals

For can and shall, the vowel shifts from /æ/ in the affirmative questions can I? and shall I? to /ɑː/ in the negative questions can’t I? and shan’t I? A vowel shift (though not the same one) also occurs going from affirmative will I? to negative won’t I?.

Now, for am, the same vowel shift occurs in going from affirmative am I? to negative aren’t I). It so happens that we spell /ɑːnt aɪ/? as aren’t I, but maybe that spelling is coincidental and misleads us into thinking this form is the same as the one for the other 4 combinations of person and number.

I haven’t explored how the form aren’t in aren’t I? is best viewed:

• morphologically as being identical with the forms of the 2^nd person plural and of all persons in the plural?
• phonologically and orthographically as a written representation of a sound change that also occurs for can and shall?   

Do English verbs in fact  inflect for number?

In a previous post, I discussed a different approach to English verbal agreement (by Richard Kayne).  Kayne suggests that:

• English verbs inflect only for number, not for person.
• only 3^rd person subjects are marked for number. A 3rd person subject is marked grammatically as singular (he / she / it) if the subject is a single person or thing but as plural (they) if the subject is more than one person or thing.
• the 1^st person (I and we) is not marked at all for number. On the other hand, 2^nd person (you) subjects are always marked grammatically as plural, even if they refer to only one person.
• the suffix -s is used when the subject is marked grammatically as singular, which occurs only for 3^rd person.
• a default form with no suffix is used for all other subjects: 3^rd person plural; 2^nd person (always marked as plural); and 1^st person (unmarked for number).

I summarised Kayne’s analysis at The English verbal ending -s – Language Miscellany

Kayne also discusses briefly—and rejects—an analysis that the English present tense inflects only for gender, but not for number or person. On that analysis, the suffix -s is required when the subject is either a pronoun displaying gender (he / she / it) or a noun (or noun phrase) that can be replaced by one of those pronouns. 

Conclusions

English seems to distinguish 3^rd person singular from all 5 other combinations of person and number in several places: verbal inflection (present tense); inflection of pronouns; and verbal inflection (in negative questions involving the copular be).

In a present tense non-modal clause, if the subject is a noun (not a pronoun), English uses the suffix -s once and once only:

• to mark the verb if the subject is singular. No suffix appears on the subject.
• to mark the subject if the subject is plural. No suffix appears on the verb.

English has 4 suffixes (3^rd person singular present tense; regular noun plural; Saxon genitive; contracted is) with almost the same sound /s/ and almost identical phonetic behaviour. Similar phonetic behaviour also occurs for the past tense weak verb suffix -d.

Follow up: one suffix or 3?

This post discusses 3 English suffixes in -s which have a similar phonetic shape and similar phonetic behaviour. Is that similarity mere coincidence? Is this one suffix with 3 different uses or are we looking at 3 different suffixes? I will discuss that in a future post looking at, for example, Berg (2011).  

Sources

Syntactic Theory: a formal introduction, Ivan A Sag, Thomas Wasow, Emily M Bender (2^nd edition, 2003), pages 109-111

Notes on English Agreement, Richard S Kayne (2000a), reprinted in Kayne (2000c) as chapter 10

Agreement and Verb Morphology in Three Varieties of English, Richard S Kayne (2000b), reprinted in Kayne (2000c) as chapter 11

Parameters and Universals, Richard S Kayne (2000c)

A diachronic frequency account of the allomorphy of some grammatical markers, Thomas Berg, Journal of Linguistics, volume 47, issue 1 (March 2011)