Irregular irregulars

Note: this post is intended for Spanish verb fiends only! Others read at your peril!

The Spanish verb system is riddled with irregular verbs, but at least they fall into discernible patterns. For example, verbs that end in -ir and have a stem change in the present tense are also irregular in the preterite, imperfect subjunctive, and gerund. These fall into three groups:

  • o/ue/u
    * Example: dormir ‘to sleep’, duermo ‘I sleep’, durmió ‘he slept’, durmiendo ‘sleeping’
  • e/ie/i
    * Example: sentir ‘to feel’, siento ‘I feel’, sintió ‘he felt’, sintiendo ‘feeling’
  • e/i/i
    * Example: servir ‘to serve’, sirvo ‘I serve’, sirvió ‘he served’, sirviendo ‘serving’

The silver lining to this cloud of complexity is that it is, at least, predictable. As implied above, there are no exceptions to this pattern, i.e. -ir verbs with a stem change in the present tense that are regular in the preterite and the gerund.

Or are there?

To my horror, and great interest, I learned just today of two exceptions: cernir ‘to sift’ and hendir ‘to slit open’. Despite their present-tense stem changes (ciernohiendo, and so on) they are regular in the preterite (cernió, cernieronhendió, hendieron), imperfect subjunctive (cerniera, hendiera, etc.), and gerund (cerniendo, hendiendo). You can see the full conjugations here and here.

Discernir and concernir share the same irregularity as cernir, as you might expect. (This is why I made sure to use the English cognate discernible at the beginning of this post. 😉 )

Not surprisingly, the Real Academia’s Diccionario panhispánico de dudas contains warnings against forms such as hindióhindieron, and cirniendo.

Fortunately, there is a logical explanation for these irregular irregulars: cernir and hendir are variants of the -er verbs cerner and hender, from Latin cernĕre and findĕre. In other words, they are innovative -ir verbs that still think they are -er‘s with respect to this irregular pattern. If I can attempt a wacky analogy, they’re akin to someone who dyed their hair but red still lacks the freckles that a natural redhead would have.

Just for fun, I used the Google ngram viewer to trace the history of cerner, cernir, hender, and hendir. None of these verbs is very common, but the -ir variants have definitely caught up to the older -er forms over the last two hundred years or so, and, in fact, have managed to surpass them.

(Post continues after graphic.)

If you look at a shorter time period, you can clearly see hender nose-diving to fall just behind hendir. It’s pretty cool.


6 thoughts on “Irregular irregulars

  1. Susan R

    This is amazing, both the irregular verb forms and also the charts on usage over time! Thanks for sharing this. I have a question that maybe you have previously addressed, but I couldn’t find it. I’m wondering how to explain the fact that ‘conjugar’ is not conjugated like a compound form of ‘jugar’, as usually (almost always/always?) happens in these compound verb examples: componer, mantener, etc.

    1. jhochberg Post author

      Jugar is, of course the only Spanish verb with a u –> ue pattern. Etymologically speaking, it should have been jogar. Please see this earlier post for Tom Lathrop’s speculation that the word changed to become less similar to joder. In that case conjugar would not have had a similar motive to change. (I apologize for getting all anthropomorphic on these verbs…)

    2. Daniel Nappo

      This is a fascinating post!

      I may be mistaken, but doesn’t jugar and conjugar come from different Latin verbs? I didn’t know what Judy said about jugar being the only u > ue verb. That’s remarkable. I think, though, that IOCARE > jugar and CONIUGARE > conjugar (IUGARE is like, put a “yoke” on, or–I guess–put a verb to work).

      With the above examples of cerner > cernir and hender > hendir, would these be examples of the differentiation that occurred with Spanish root vowels? The explanation I’m thinking of would be Penny’s History 2nd ed (p. 157).

      1. jhochberg Post author

        Indeed, jugar and conjugar are not related. A good point!
        I just reread that page from Penny and don’t think it relates to the evolution of cerner > cernir and hender > hendir. In fact, he says that –ir verbs are less likely than –er verbs to have the vowel /e/ in the root.
        Rather, it was probably part of a larger tendency for –er verbs to turn into –ir verbs. He discusses this on page 173. Other –ir verbs that used to be –er verbs include herir, cumplir, pudrir, huir, morir, and recibir.

  2. Daniel Nappo

    Thanks, Judy! I’m still trying to get my head around the root vowels in verbs. It’s confusing because I’m not sure if it’s applicable to both the infinitive forms and the preterit forms. I came across a term in an English language history–an ablaut–and I was thinking that it was perhaps related. Given that the language is predominantly spoken, the differentiation occurred to distinguish the mid vowels from each other. I’ll keep at it! Great post!!

    1. jhochberg Post author

      Ablaut is one of those terms from graduate school that I always think I should understand, but don’t…figure that if I ever have to become an ablaut expert, I will! However, what you said about distinguishing mid-vowels reminded me of the Siglo de Oro consonant shifts where the /x/ of ojo developed (nice and far back in the throat) and likewise the ‘th’ of Castilian cerveza — nice and forward in the mouth — from consonants that were originally much closer together, with the result of differentiating sounds that were too similar.


Leave a Reply

Your email address will not be published. Required fields are marked *