Compared to English, Spanish is very consistent within its rules about verbs. The endings for the three kinds of verbs—grouped as -ar, -er, and -ir verbs—are pretty consistently regular, and few words that break the regular pattern don’t follow another one.

I know that some verbs, e.g. ir, are weird because they come from multiple Latin words. Wiktionary says this about ir:

The infinitive and forms beginning with i or y are from Latin īre, present active infinitive of ...the forms beginning with v from corresponding forms of vādō; the forms beginning with f from forms of sum.

So, in this case, the weirdness was created after Spanish started to break away from Latin, and thus not something from the older language.

How consistent are verbs in ancient Latin, in general, compared to in Spanish? Have any studies gotten good statistics on how regular verbs were in it?

  • 2
    Spanish ir is a suppleted verb. There were a few of these in Latin, but not many; fero-ferre-tuli-latus, sum-esse-fui are the ones that come to mind. – Bert Barrois Nov 10 '19 at 12:19

Significantly less so.

One way to measure the "regularity" of a language's verbs is, how much information do you need to memorize for each verb in order to conjugate it properly?

For Spanish, you don't need very much. If you have the infinitive, you can deduce the rest from that. There are a few irregular verbs, but those are a relatively small fraction of the total.

In Latin, you generally need to know four separate forms, the "principal parts", which provide information you can't get from the infinitive. For example, there are a few different ways to form the perfect (basically simple past) tense stem from the present stem: sometimes you add a -v- (amā-amāv- "love"), sometimes you add an -s- (dīc-dīx- "say"), sometimes you lengthen the vowel (ven-vēn- "come"), sometimes you reduplicate the first syllable (curr-cu-curr- "run"), and so on. There are a few general trends, like how most stems ending in -ā- add a -v-, but they're more like guidelines than hard rules. So you need to memorize the past stem alongside the present stem. And then you can usually figure out the supine stem from the past stem, but not always, so you need to memorize that too, and you can usually figure out the present stem from the infinitive stem, but not always, so you need to memorize that too, and in the end you're learning four separate forms for each verb.


I know that some verbs, e.g. ir, are weird because they come from multiple Latin words.

This is called suppletion, and happens in Latin as well as Spanish. (In fact, it happens in most inflecting languages—consider English "go" vs "went", or "be" vs "is" vs "was".) But suppletive verbs are generally a small minority, so they don't matter as much in terms of the overall regularity of the language.

| improve this answer | |
  • I have never heard that form called the "aorist" or "simple past". In my experience it is always referred to as the "perfect", even though it does not have the specific meaning that perfect forms have in some languages such as English. – Colin Fine Nov 10 '19 at 19:45
  • @ColinFine One of my professors was a Greek specialist, so he probably just applied the Greek term to the Latin form. (I've kept using it mostly because it lets me distinguish between the two different tense-aspect combinations with the same form, which is sometimes very useful.) – Draconis Nov 10 '19 at 19:46
  • 1
    @ColinFine (Fixed, though) – Draconis Nov 10 '19 at 20:21
  • As a counterpoint, though, the principal parts are almost the only locus of irregularity in Latin verbs -- there are only a handful of verbs that inflect irregularly in other ways, while Spanish has many more (en.wikipedia.org/wiki/Spanish_irregular_verbs). – TKR Nov 11 '19 at 23:19

I'll compensate for my ignorance of Latin by claiming we cannot answer this with something like "42% more irregular".

How would we measure this?

Would we weight the verbs by their frequency? In which corpus? Would we count very rarely used tenses? What if the irregular falls into a pattern shared with other verbs? What if it's predictable by phonetics and not an actual stem change? What if one language reflects it in pronunciation but not in orthography? What if it varies by dialect?

I hold: we could invent some statistic to support either conclusion. People graduate and publish with far more subjective scholarship.

Is ir a representative sample?

It's hard to theorise about a language-wide shift based on the example of one verb that is moderately to highly irregular in almost every language for which I know the conjugations - English, Russian, Armenian, Spanish, Italian, French, Serbo-Croatian, Greek, German, Yiddish, Alemannic - despite a tendency to regularise verbs... - and even - consulting Wiktionary... - Hungarian. Especially in Romance and SAE.

Across languages, to go is usually as irregular as any verb after to be, along with to come and modal verbs. In fact, Which languages have a conjugation of to go that is totally regular? is a good question. (The best candidates that come to my mind are maybe Macedonian иде and Turkish gitmek and literary Persian رفتن. But all with caveats.)

There are forces that lead to regularisation of verbs, just like their are forces that lead to irregular verbs, including suppleted verbs, and generally to verbal morphology in the first place. Regularity at the lexical level is a function of frequency, regularity at the language level is a function of typology. So it's really a language change cycle question.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.