Better formulations of verb conjugation patterns

[samfrances]

Does anyone get the feeling that we can do a better job of formulating grammatical patterns than the traditional textbooks do? For example, Spanish present tense usually lists three different sets of endings:

-ar (e.g. hablar)

-o
-as
-a
-amos
-áis
-án

-er (e.g. comer)

-o
-es
-e
-emos
-éis
-en

-ir (e.g. vivir)

-o
-es
-e
-imos
-ís
-en

But there are so many commonalities and patterns here. It feels like you could make it easier to remember by splitting the endings into two. So verb form = stem + "theme vowel" + ending (despite my rusty old linguistics degree, I don't have better words for these)

Theme vowel (Ø means absence of a vowel or a "null" vowel):

- Ø
- a / e
- a / e
- a / e / i
- á / é / Ø
- a / e

"Ending" is common for all verb classes:

- o
- s
- Ø
- mos
- is
- n

The only thing this doesn't completely describe is the emphasis of "vivís", but it seems a lot more concise than the traditional description. And the emphasis patterns are the same for all verb classes, so perhaps you could describe them separately, or describe the "null" theme vowel in the vosotros form as having an emphasis that "spreads" to the "is" in the ending.

Has anyone come across successful attempts to call out these sorts of patterns in a better way than the traditional textbooks? The only thing I have found is the attempt to describe spanish verbs in terms of "principal parts" in Spanish Irregular Verbs Up Close, but that doesn't seem perfect either.

I'm sure there's a lot of linguistic research into why these verbs behave the way they do etc., but I'm talking more about how we describe them for the purposes of teaching / remembering them than the "true linguistic theory" about them.

[Iversen]

I make 'green sheets' when I start learning a new language - but since I learned Spanish long before I got that idea I don't have any sheets for Spanish - sorry. However I do think that it is sensible to isolate endings from whatever is before them (minus stems and prefixes). In some languages the linguists do operate with conjugations based on stem vowels (or lack thereof, called consonant stems), but the nice logic of this is lost when an ordinary language learner can't use that system to anyhing. I have read about a -j stem somewhere (was it Polish?) that defined a conjugation that simply doesn't have any j's anymore. That's silly, but there are so many languages where the conjugations differ on just the vowel before the endings - and then maybe one or two sets of possible endings.

I once returned from the Philippines and then I tried to make sense of the Tagalog verbs - but never got around learning the language itself. However the verbs have things like reduplications, infixes and something that looks like endings. And then the books I used warn learners not to think about the forms in terms of Westernese tempora and modi, but they weren't very explicit about what you could use instead - except that the system operated with something you could call conjugations. Nevermind, that doesn't matter. The point is that I made the green sheet below where the different kinds of elements are marked in different colours:

Philippine Verbs.jpg

As for Spanish there are some vowel changes that depend on the place of the accent, but not really on the conjugation - so these wowel changes should be listed separately, not with all the possible endings for each variation on the theme. This principle is especially important for the Celtic languages where there are changes at both ends at the main words, but those at the beginning of words are based on preceding words (particles, prepositions..) and things like gender (of substantives), and they have next to nothing to do with whatever happens at the other end of the word. So evidently the complexity of the tables skyrockets if you try to illustrate all possible combinations, and the solution is obviously not to do so - keep separate things separated.

In grammar books the conjugations are mostly defined by referenceds to thematic vowels (a heritage from Latin grammarians), but in one language there was a reference to -je stems, despite the fact that the j's weren't visible to the language learners anymore. That's stupid. All categories should be defined by things that can be seen even by a casual learner. Besides the conjugations are normally treated separately, one after the other, but then you can't make easy comparisons, so the main idea behind my sheets is to show the endings with conjugation-defining preceding wowels on a single sheet. The price is that for obvious space reasons I can't illustrate the whole thing with concrete words, and I also have to make a special sheet for the irregular verbs. But actually, for Latin there is an excellent system where you just quote four important forms (three for deponent verbs), and once you know these forms for a given verb the formation of the other forms in a certain part of the table system are easy to construct.

I also show my verbal table for Icelandic (unusual insofar it uses a concrete word as illustration). In all Germanic languages there are strong verbs with wowel changes, but not much in the way of endings, and weak verbs without wowel change, but at least in Icelandic and High German a full set of person-and-number endings. So the main thing to know about 'strong' verbs is how the stem wowel changes, and I have shown that for Icelandic at the upper right - isolated from the table of endings. As for Spanish the changes that only concern a stem vowel - based on the position of the stress - don't really make a verb irregular. At most they become semiirregular, but you could just as well say that such verbs are semiregular, just exploiting a predictable vowel shift. However a verb like "ser" in Spanish is obviously as irregular as can be.

Here Icelandic:

Icelandic Verbs.jpg

And now Latin:

Latin Verbs.jpg

The few genuinely irregular verbs are those where this isn't possible (like "esse"), and there you have to the learn the forms one by one. So the table with the endings just has to be supplemented by primo: a list of the main forms of the more or less irregular verbs, secundo: all forms of the really irregular ones. As you can see I have simplified the notation of compound forms. It is absolutely idiotic to quote all forms of a compound form when the only things that differ are the normal endings of the participle (according to rules known from the adjectives) - the forms of the auxiliary verb are the usual ones.

Which actually takes us to the past tenses in the Slavic languages. There you find the compounds with a Past Passive participle plus some form of 'to have' which also are used in English, but half of them have past tenses that are based on a Past ACTIVE participle which isn't found in the Germanic and Romance languages. I have shown a sketch for the Serbian sheet below, and here you can see the adjective-like particle "imao/imala/imalo - imali" in combination with forms of the 'to have' verb - the present, the imperfect and the remnant of an aorist (not much used in isolation in Serbian, but a very real and alive terror for language learners in Bulgarian), and then it is clear what happened in Polish and Russian: the auxiliary simply went down the drain and disappeared, leaving just the participle. In Polish they added a set of pronominal endings to this stub, which makes the Polish system much more complicated.

Serbian verb.jpg

Bulgarian Verbs.jpg

NB: the sheet above hasn't been completed - I have reorganized the system later and written about the process in this thread, but not yet made the corresponding sheet. One item more on my 'to do'-list..

And now: how does this help the green sheet developer? Well, if you have a sheet for short adjectives in Russian then you recognize the endings of the 'past' tense, but since they are so few in number you can just as well quote them. In the range from Czech and Slovak down to Macedonian and Bulgarian you have to show the forms, but since all the auxiliary verb forms are listed on a second sheet and the adjectival endings of the participle part are, well, the usual adjectival ones, then it suffice to show each 'tense' with one example on one line (or maybe as 1.person singular of the auxiliary and the participle endings on the compact form zero/a/o//i ). And the names af those tenses are in most languages so irrelevant bordering on idiotic that you just choose one set of names and use that for all your sheets - better be consistent than bound to confusing old traditions.

[chove]

Don't know if it's what you mean but I always write Spanish verbs as two lists with singular next to plural. How does this help? Well, fui looks a bit like fuimos, and fue goes with fueran, so I can see the patterns a bit more easily.

[samfrances]

Don't know if it's what you mean but I always write Spanish verbs as two lists with singular next to plural. How does this help? Well, fui looks a bit like fuimos, and fue goes with fueran, so I can see the patterns a bit more easily.

That sort of thing, yes.

Or for another example, a professor called Eric Vogt has a system where you can unify the rules for regular and irregular verbs and treat them all as regular.

The trick is to treat verbs as having more "principal parts", rather than just the infinitive, and memorize them. (A system similar to that used for latin, I believe).

For example, if you say that verbs have 3 principal parts - the first person singular (yo) form, the second person singular (tu) form, and the infinitive, you can eliminate most irregularity from the present tense.

e.g. "hablo", "hablas", "hablar"
e.g. "tengo", "tienes", "tener"

You can derive a "tu stem" from the memorized tu form by knocking off the -as or -es. So the "tu stem" for "hablar" is the same as the infinitive stem, "habl", but the "tu stem" for "tengo" is "tien".

Then you conjugate as follows:

yo: memorized principal part
tu: memorized principal part
el: tu stem + ending
nosotros: infinitive stem + ending
vosotros: infinitive stem + ending
ellos: tu stem + ending

So for the verb "hablar" (which we can now refer to in full as "hablo", "hablas", "hablar")

hablo
hablas
habl + a
habl + amos
habl + áis
habl + an

And "tener" (which we can now refer to as "tengo", "tienes", "tener"):

tengo
tienes
tien + e
ten + emos
ten + éis
tien + en

And just like that, most of the common irregularities in the present tense fall away. You can conjugate "hablar" and "tener" according to exactly the same pattern.

Vogt extends this to deal with other tenses, so that verbs end up with:

six forms of any verb that you need to know in order to master this system: (1) the first- and (2) second-persons singular of the present indicative; (3) the infinitive; (4) the first-person singular of the preterit indicative; and the two participles— (5) the gerund and (6) the past, or passive, participle.

e.g. tengo, tienes; tener; tuve; teniendo, tenido

You can't eliminate true irregulars like "ir", but a whole lot of irregularity gets eliminated this way.

Sorry for geeking out, but I just find this so beautiful! I just wonder how many more opportunities there are for reformulating the grammar rules to make things easier, more regular etc.!

[samfrances]

I make 'green sheets' when I start learning a new language - but since I learned Spanish long before I got that idea I don't have any sheets for Spanish - sorry..

That's super-interesting, although it will take a while to digest. Do you write more about the "green sheet" idea elsewhere?

[jeff_lindqvist]

Here you are:
3.3. Green sheets

[tungemål]

You have a good point, but I don't think this is a good idea for learning a language.

The deductive way: going from defined rules -> deducing grammatical forms -> using the language -- is probably not how the brain works best. The inductive way is probably better:

Language examples -> inducing grammatical forms (but also studying grammatical forms -> noticing the rules yourself (after you've learned the conjugations).

I stopped writing down endings - I always write down the full word. With an example sentence is probably best.

[samfrances]

You have a good point, but I don't think this is a good idea for learning a language.

The deductive way: going from defined rules -> deducing grammatical forms -> using the language -- is probably not how the brain works best. The inductive way is probably better:

Language examples -> inducing grammatical forms (but also studying grammatical forms -> noticing the rules yourself (after you've learned the conjugations).

I stopped writing down endings - I always write down the full word. With an example sentence is probably best.

I don't know, there's definitely deductive rules I find helpful.

For example, past subjunctive = ellos preterite form with the -on replaced with -a.

e.g. tener -> tuvieron -> tuviera

Obviously I'm not consciously doing this with verbs I use often in this form. But with new or less used verbs, its useful to have this in your back pocket.