While studying bases from Latin oblique stems recently, I came to reevaluate the “stem suffix.” I had previously identified forms like the -it in addition as a stem suffix. English addition comes from the oblique stem additiōn- of the Latin additiō (genitive additiōnis), which comes from additus + -tiō, which is the past participle of the verb addere, which comes from ad- + -dere. However, upon further investigation, I realized that the -it comes from -ītus, the Latin past participle suffix.
The Latin -ītus developed into both -it and -ite in English as in <Ex + Plice + it → explicit> and <op + Pose + ite → opposite>. I thus represent the suffix as the parenthetical -it(e) to account for the allomorphs -it and -ite in the same morpheme entry. The parenthetical -it(e) is also important for a word like addition because the form of the -it(e) suffix is unknowable. Both <ad + D + it + ion → addition> and <ad + D + ite + ion → addition> are accurate word sums. The <t> of -it does not double because letters in affixes never double, and the <e> of -ite is replaced by the vowel-initial -ion.
The English suffix -it(e), which surfaces as -it and -ite, ultimately comes from the Latin -ītus. The Latin -ītus forms adjectives and perfect passive participles, primarily of fourth conjugation verbs (-īre) and sometimes of second conjugation verbs (-ēre). In Latin, the -ītus sometimes surfaces as -īt- when followed by another suffix as in Latin additiō, which comes from additus + -iō. Thus, the English -it(e) comes from Latin -īt- and -ītus, with the -īt- coming from -ītus. In other words, the English -it(e) ultimately comes from Latin -ītus.
The English -it(e) from Latin -ītus forms nouns, adjectives, and verbs. Examples of the noun suffix are <Fru + it(e) → fruit>, <Fru + it(e) + ion → fruition>, and <Habe + it(e) → habit>. English fruit comes from Latin fruitus, a past participle of the verb fruī, via Old French fruit. English fruition comes from Latin fruitiō, which comes from fruitus + -iō. English habit comes from Latin habitus, which is the past participle of the verb habēre. Examples of the adjective suffix are <Fin(e) + it(e) → finite> and <Tace + it(e) → tacit>. English finite comes from Latin fīnītus, the past participle of the verb fīnīre. English tacit comes from Latin tacitus, the past participle of the verb tacēre. Examples of the verb suffix are <e + D + it(e) → edit> and <Un(e) + it(e) → unite>. English edit ultimately comes from Latin ēditus, the past participle of the verb ēdere. English unite comes from Latin ūnītus, the past participle of the verb ūnīre.
English does have another -it(e) suffix that comes from Old French -ite from Latin –ītēs from Ancient Greek –ítēs. The second -it(e) suffix is a noun and adjective suffix, demonymic suffix, agent suffix, chemical suffix, and geologic suffix. Example of this -it(e) suffix include <Calc + it(e) → calcite>, <Ludd + it(e) → luddite>, and <Palm + it(e) + Ole + ic → palmitoleic>. English also has an <It(e)> bound base from Latin itus, the past participle of the verb īre, as in <Ex + It(e) → exit> and <Amb + It(e) + ion → ambition>.
In addition to the pronominal <It> free base as in <It → it> and <It + Self → itself>, English also has two -it suffixes. The first -it suffix is the frequentive and intensive verb suffix from Latin -itāre. Examples include <In + Habe + it → inhabit> and <Vome + it → vomit>. English inhabit comes from Latin inhabitāre, from in- + habitāre, from habēre + -itāre. English vomit comes from Latin vomitāre, from vomere + -itāre. Other word families with the frequentive and intensive -it suffix include agitate from Latin agitāre from agere + -itāre, hesitate from Latin haesitāre from haesum + -itāre from haerēre, and visit from Latin vīsitāre from vīsere + -itāre, among others. The other -it suffix is the rare diminutive in <Gamb + it → gambit>, which is the Anglicized form of the Italian -etto from Late Latin –ettus, –ittus.
All the forms of -it and -ite etymologically come from an identifiable classical suffix: Latin -ītus for noun/adjective/verb -it(e), Ancient Greek –ítēs for noun/adjective -it(e), and Latin -itāre for frequentive and intensive -it. No -it and -ite comes from a classical stem. Thus, none of the above identified -it, -ite, or -it(e) suffixes are stem suffixes.
What about the words limit, termite, concomitant, equites, military, regurgitate, and stipitiform? Is the -it or -ite a separate morpheme? The short answer is no. The long answer requires delving into classical oblique stems.
The English word limit comes from the Latin līmes, which is a third declension noun. The Latin third declension refers to the large and diverse group of nouns and adjectives identified by the genitive singular -is ending. But English limit does not come directly from the nominative singular līmes. Instead, English acquired limit from the oblique stem līmit-. Many Latin third declension words have a nominative singular stem that differs from the oblique stem (cases and numbers other than the nominative singular and sometimes the vocative singular). The oblique stems are the older forms, preserving the historical origin, while the nominative singular underwent various changes. For example, with the noun līmes, some of the oblique forms include the nominative plural līmitēs, genitive singular līmitis, and accusative singular līmitem. The endings are the nominative plural -ēs, genitive singular -is, and accusative singular -em, which attach to the oblique stem līmit-. Thus, the <it> in the English limit is part of the free base from the Latin oblique stem līmit-, not a separate morpheme.
The -it or -ite in termite, concomitant, equites, military, regurgitate, and stipitiform is also part of the base, not a separate morpheme. <Termite> ultimately comes from the oblique stem termit- of Latin termes (genitive termitis). The base of <con + Comite + ant → concomitant> comes from the oblique stem comit- of Latin comes (genitive comitis). The base of <Equite + es → equites> comes from the oblique stem equit- of Latin eques (genitive equitis). The base of <Milite + ary → military> comes from the oblique stem mīlit- of Latin mīles (genitive mīlitis). The base of <re + Gurgite + ate → regurgitate> comes from Latin -gurgitāre from the oblique stem gurgit- of gurges (genitive gurgitis). The base of <Stipite + i + Form → stipitiform> comes from the oblique stem stīpit– of Latin stīpes (genitive stīpitis). In all these bases, the -it or -ite is part of the bases because the sequence of graphemes is part of the Latin oblique stem, not a separate morpheme.
Also note that English often developed two separated but related bases from the Latin third declension. One base comes from the nominative singular and the other from the oblique stem. For example, the bound base <Lime> and free base <Limit> come from Latin līmes (genitive līmitis) as in <Lime + es → limes> and <Limit + ed → limited>. The bound base <Term> and the free base <Termite> come from Latin termes (genitive termitis) as in <Term + es → termes> and <Termite + i + Cide → termiticide>. The bound bases <Come> and <Comite> come from Latin comes (genitive comitis) as in <Come + es → comes> and <Comite + al → comital>. The bound bases <Equ> and <Equite> come from Latin eques (genitive equitis) as in <Equ + ine → equine> and <Equite + es → equites>. The free base <Gurge> and the bound base <Gurgite> come from Latin gurges (genitive gurgitis) as in <Gurge + ed → gurged> and <re + Gurgite + ate → regurgitate>. The free base <Stipe> and the bound base <Stipite> come from Latin stīpes (genitive stīpitis) as in <Stipe + Less → stipeless> and <Stipite + ate → stipitate>.
To play devil’s advocate for the so-called stem suffix -it, the -it appears to attach to the base from the nominative singular of the third declension. For example, the bound base <Lime> denoting “border, path” from Latin līmes appears to take the -it to form the word limit: <Lime + it → limit>. I define the morpheme as “the smallest meaningful linguistic unit of a language.” If the -it were a suffix, then the stem suffix would be akin to a connecting vowel, connecting two morphemes but not really expressing semantic meaning. The “meaningfulness” of the stem suffix -it would be “connecting.”
But not all classical oblique stems end in -it. For example, the oblique stems of Latin mātrīx (genitive mātrīcis) and Latin index (genitive indicis) are mātrīc- and indic-, which end in -ic. Either <Mat(e)r> (with the word mater as evidence of the parenthetical <e>) and <Matrice> both come from Latin mātrīx (oblique stem mātrīc-) and <D> and <Dice> come from Latin index (oblique stem indic-), or the -ic is a stem suffix. The oblique stem of Latin proboscis (plural proboscidēs) (no attested genitive form) is proboscid-, which ends in -id. Either <Bosc> and <Boscide> both come from Latin proboscis (oblique stem proboscid-), or the -id is a stem suffix. The oblique stem of Latin palūs (genitive palūdis) is palūd-, which ends in -ud. Either <Pale> and <Palude> both come from Latin palūs (oblique stem palūd-), or the -ud is a stem suffix. The oblique stem of Latin abiēs (genitive abietis) is abiet-, which ends in -et. Either <Abi> and <Abiete> both come from Latin abiēs (oblique stem abiet-), or the -et is a stem suffix. The oblique stem of Latin corpus (genitive corporis) is corpor-, which ends in -or. Either <Corp> and <Corpor> both come from Latin corpus (oblique stem corpor-), or the -or is a stem suffix. The oblique stem of Latin genus (genitive generis) is gener-, which ends in -er. Either <Gene> and <Gener> both come from Latin genus (oblique stem gener-), or the -er is a stem suffix. The oblique stem of Latin coccȳx (plural coccȳgēs) (no attested genitive form) is coccȳg-, which ends in -ȳg. Either <Cocc> and <Coccyge> both come from Latin coccȳx (oblique stem coccȳg-), or the -yg is a stem suffix. The oblique stem of Latin calyx (genitive calycis) is calyc-, which ends in -yc. Either <Caly> and <Calyce> both come from Latin calyx (oblique stem calyc-), or the -yc is a stem suffix. The oblique stem of Latin nōmen (genitive nōminis) is nomin-, which ends in -in. Either <Nome> and <Nomine> both come from Latin nōmen (oblique stem nōmin-), or the -in is a stem suffix. The oblique stem of Ancient Greek stóma (genitive stómatos) is stómat-, which ends in -at. Either <Stome> and <Stomate> both come from Ancient Greek stóma (oblique stem stómate), or the -at is a stem suffix.
Word endings in languages of origin often become suffixes in English. For example, the singular noun ending -is of Latin crisis and basis became the singular noun suffix -is of crisis and basis in English. The corresponding plural noun ending -ēs of Latin crisēs and basēs become the plural noun suffix -es of crises and bases in English. (Note that the Latin plural -es of crises and bases is not the same as the English plural -es of babies and couches. The two plural suffixes are homographic but not homophonic.) The singular noun ending -a of Ancient Greek skhêma and pneûma became the singular noun suffix of schema and pneuma in English. The corresponding plural noun ending -ata of Ancient Greek skhḗmata and pneúmata became the plural noun suffix of schemata and pneumata in English.
An argument against stem suffixes in English from classical oblique stems is that the final letters on the oblique stem were never an affix in Latin or Ancient Greek. The -ic, -id, -ud, -et, -or, -er, -yg, -yc, -in, and -at are part of the classical oblique stem. To reiterate, the oblique stems of the third declension preserve the historical origins of the words. The nominative singular underwent various changes, and the oblique stems are the older forms. The nominative singular developed from the oblique forms. Essentially, nominative singular lost the final syllable preserved on the oblique stem. The oblique stem did not gain a morpheme.
For a morpheme to exist in English, the morpheme requires morphological, etymological, and semantic evidence. While sequences of graphemes potentially identified as stem suffixes are etymologically attested in Latin and Ancient Greek as part of the oblique stem of the third declension, the hypothesized morphemes lack morphological evidence. The final letters were always part of the oblique stem, not a separate morpheme. The semantic evidence is also questionable. An affix akin to a connecting vowel would suffice. However, the lack of morphological evidence disproves the hypothetical stem suffix. The oblique stems in Latin and Ancient Greek did not end in a suffix, so neither do the English words. Instead, the classical third declension usually developed into two separate but related bases in English.
Positing a stem suffix would make certain word sums “work.” For example, positing an -ic suffix would put singular matrix and plural matrices in the same <Matr> word family: <Matr + ix → matrix> and *<Matr + ic + es → matrices>. But one cannot invent a morpheme simply to make a word sum work. The morphological, etymological, and semantic evidence must support the existence of a morpheme.
Even though sequences of graphemes like -ic, -id, -ud, -et, -or, -er, -yg, -yc, -in, and -at occur at the end of oblique stems in Latin and Ancient Greek, the graphemes belong to the oblique stem in the source language and subsequently a base in English. Just as English developed two separate but related bases from the second and fourth principal parts of some Latin verbs, English likewise often developed two separate but related bases from the nominative singular and oblique stem of some classical third declension Latin and Ancient Greek nouns and adjectives. The oblique stem base analysis handles classical bases from the third declension elegantly with a single consistent principle rather than requiring an ever-expanding inventory of morphologically-unsupported stem suffixes.
Additionally, the final sequence of graphemes on classical oblique stems do not behave morphologically like the connecting vowels -e-, -i-, -o-, and -u-. For example, the Hellenic -o- freely connected many bases in Ancient Greek to form compounds and continues to function connectively in Modern English. English geography comes from Ancient Greek geōgraphía. The -o- between the <Ge> and the <Graph> is a connecting vowel. I created the word Morphodex in Modern English using the connecting vowel -o- between <Morph> and <Dex>. While productivity is not a requirement for the existence of a morpheme, the posited “stem suffix” is etymologically bound. The final sequence of <ice> occurs in <Matrice> because the oblique stem of Latin mātrīx is mātrīc- rather than an -ic affixing to matr- to form mātrīc-.
Rebracketing, or morphological reanalysis, is a linguistic phenomenon in which the boundaries between words or morphemes shift over time. For example, English apron comes from a rebracketing of a napron, with English speakers misinterpreting the <n> as part of the determiner an rather than the initial letter of napron. Many suffixes of Latin origin that begin with <a> were rebracketed from the -a- of the stem of first conjugation Latin verbs plus another suffix. While rebracketing can form new morphemes, English borrowed the oblique stem mātrīc- in the form of the bound base <Matrice> as a complete unit. Rebracketing mātrīc- into <Matr> and a hypothesized -ic does not produce a suffix in the same way as the rebracketing of the -a- of the stem of first conjugation Latin verbs produced suffixes like -able, -ability, and -ac(u)le. The posited -ic is bound to the word family from Latin matrix but not freely affixing to other bases.
Although the posited stem suffix appears morphologically similar to connecting vowels, all examples are better explained as descendants of identifiable suffixes or the oblique stem base analysis. As previously stated, one cannot invent a morpheme simply to make a word sum work. For example, in a word like <ad + D + it(e) + ion>, the -it(e) ultimately comes from the Latin -ītus suffix. In the word limit, the final <it> is part of the base, which English borrowed from the oblique stem of the third declension Latin līmes (genitive līmitis). Understanding that English often developed two different but related bases from the nominative singular and oblique stem of some Latin and Ancient Greek words—rather than positing multiple etymologically-conditioned but morphologically- unsupported stem suffixes—provides the most elegant morphological analysis.
