The SAE Grammar Test

[HoskhMatriarch]

A while back there was the SAE (Phonology) Test. However, the SAE Sprachbund was initially defined primarily, if not exclusively, by grammatical features, and anyways, I get bored as hell of talking about phonology all the time when we can talk about syntax, morphology, pragmatics, and all the fun stuff (what inspired me to make a conlang was pretty much my love of grammar, as for a while the phonology in my favorite conlang at the time was pretty much just a German dialect with a few sounds added and taken away because I didn't GAF about the phonology, but the grammar was most certainly not). I'm also not going to title this "how not make an non-SAE grammar" or something because who cares if your grammar is SAE if that's what you want. This test was made by someone on the CWS Chat after I took the "SAE Test" and was like "where the hell are the grammar questions? It isn't titled 'SAE Phonology Test' so I expected lots and lots of grammar questions :/" (I even renamed the "SAE Test" to the "SAE Phonology Test" for a while when no one was looking but then changed it back so people wouldn't kill me if they saw it, and I might change it back too, because it is the SAE Phonology Test. Nothing wrong with a phonology test, but there's way more to a language than that).

Anyways, here is the SAE Grammar Test aka the real one:

https://docs.google.com/spreadsheets/d/ ... edit#gid=0

[Ser]

I think question 6 could be worded much better:

6. Anticausitive prominence: the intransitive verb is derived from the transitive. (e.g. The flame melts the ice -> The ice melts) [full marks for over 70% of intransitives derived from transitive; half marks for over 50%]

It's spelled "anticausative", but besides that, in Indo-European most intransitive verbs are NOT derived from the transitive. Haspelmath was specifically referring to what he calls "inchoative-causative alternations", like "boil", "melt", "get lost/lose", "rise/raise". The study Haspelmath refers to (which he himself carried out) compared only 31 such alternations. I happen to have access to the study, and the 31 pairs were:

1. wake up
2. break
3. burn
4. die/kill
5. open
6. close
7. begin
8. learn/teach
9. gather
10. spread
11. sink
12. change
13. melt
14. be destroyed/destroy
15. get lost/lose
16. develop
17. connect
18. boil
19. rock
20. go out/put out
21. rise/raise
22. finish
23. turn
24. roll
25. freeze
26. dissolve
27. fill
28. improve
29. dry
30. split
31. stop

It'd be good if this list could be provided, so that if a language has anticausatives for > 50% or > 70% of the verbs of this list they could appropriately receive some points for SAE-ness.



I have similar criticism regarding the wording of question 10.

10. Equative constructions based on an adverbial-relative clause structure. (e.g. French grand comme un élephant).

It's spelled éléphant, but besides that, it's just confusing. But this is not necessarily you guys' fault as it is of the Wikipedian who added this to the article on SAE. (EDIT: I have edited the Wikipedia article improving its wording.)

The construction Haspelmath refers to is actually French si grand comme un éléphant, so, the equivalent of "[is/are] as big as an elephant", i.e. comparisons of equality, not the equivalent of "big like an elephant". Haspelmath refers to French comme, German wie (from so...vie), Russian kak (from tak zhe X kak Y), etc., as "adverbial relative pronouns", presumably because these are related to adverbial pronouns (notably interrogative ones?). (No idea why he calls them "relative" though.)

A much better wording would be, for example:
"10. The construction for comparison of equality is based on what Haspelmath calls adverbial relative pronouns, such as the equivalent of 'how'. (E.g. French si grand comme un éléphant, meaning "as big as an elephant", literally "so big how an elephant")."



I found question 24 confusing. What is a "specific" neither-nor construction exactly? What would be an example of a "non-specific" neither-nor construction?

[Yng]

we already have one of these

viewtopic.php?f=4&t=42393&hilit=SAE+grammar

[Curlyjimsam]

I think question 6 could be worded much better:

6. Anticausitive prominence: the intransitive verb is derived from the transitive. (e.g. The flame melts the ice -> The ice melts) [full marks for over 70% of intransitives derived from transitive; half marks for over 50%]

It's spelled "anticausative", but besides that, in Indo-European most intransitive verbs are NOT derived from the transitive. Haspelmath was specifically referring to what he calls "inchoative-causative alternations", like "boil", "melt", "get lost/lose", "rise/raise". The study Haspelmath refers to (which he himself carried out) compared only 31 such alternations.

The English example is misleading: it's not clear, in English, if the transitive is derived from the intransitive or vice versa or neither - both intransitive and transitive alternants have the same form. Haspelmath is interested in pairs where the intransitive alternant is morphosyntactically marked in relation to the transitive: e.g. if you have transitive boil but intransitive ANTICAUS-boil. This contrasts with the more usual pattern outside Europe where you'd have something like intransitive boil and transitive CAUS-boil.

[mèþru]

The SAE Phonology Test is really useful for people who are still working out their grammar. I feel like most people who have enough information about their language to take a grammar test can also take a phonology test. There should really be an Overall SAE-ness Test.

[Soap]

Thank you for posting that page. I originally came to ask a different question, but I see the qnswer now. Still, Im curious about something else, if any one knows. One of the menuitems says:

7. External possessors in dative case. (e.g. German Die mutter wusch dem Kinde die Haare, lit. the mother washes the child the hair)

What would the meaning be if "Kinde" was in the genitive instead? Would it just be ungrammatical and have no meaning? This question goes for any language, not just German.

I had never efven considered such a setup. Currently, in my main conlangs (all of them have simialr case setups), using the genitive here would be the appropriate choice, and if the dative* were used it would be "the mother washes the hair for the child" ... e.g. perhaps this kid is studsying to be a hairstylist but she outsources the job of actually washing people's hair to her mom. An odd supposition, yes, but I just stuck with the sentence provided.

-----

*My conlangs dont have a dative case, jsut a dative construction, but it behaves as if it was a case.

[Ser]

7. External possessors in dative case. (e.g. German Die mutter wusch dem Kinde die Haare, lit. the mother washes the child the hair)

What would the meaning be if "Kinde" was in the genitive instead? Would it just be ungrammatical and have no meaning? This question goes for any language, not just German.

I don't know about German, but Spanish uses external possessors as indirect objects too: La madre le lava el pelo al niño. Using an internal possessor, la madre lava el pelo del niño, wouldn't be ungrammatical IMO, but it's not as common or idiomatic, and IMO it'd tend to imply a different meaning (namely that the child is not personally affected by the washing, perhaps because the hair is not attached to his head (say, maybe the mother had her child's hair cut and is now washing it for some purpose)).

Hmmm... Let's try to look at another example where the above interpretation is more difficult.

• ¡Maestra, Juan me jaló el pelo!
  'Ms. [name], Juan pulled my hair!' (said by a girl with a ponytail)

If we changed this to an internal possessor, Maestra, Juan jaló mi pelo... it'd still mean the hair is not attached to her head, or that she's not as equally affected by the hair pulling in some way like that.

[cromulant]

Nothing wrong with a phonology test, but there's way more to a language than that).

mind

blown

[Salmoneus]

Nothing wrong with a phonology test, but there's way more to a language than that).

mind

blown

We can all be glad we have Hoskh around to share these breakthroughs with us.

[Qwynegold]

I don't understand how this document works. The data that's already there, is it from various conlangers who have added it themselves?

[cromulant]

I don't understand how this document works. The data that's already there, is it from various conlangers who have added it themselves?

Look at the language names bro, those are obviously conlangs. You can spot em a mile away. Also check out the comments below the data, how "Pach'o is getting particles or something now." It is unusual for such comments to be made about natlangs.

[mèþru]

I added Khasoítoí, which is my conlang.

[garysk]

Sorry for bumping this thread, and sorry for my ignorance. In my ken, SAE is an acronym for the Society of Automotive Engineers, which is commonly seen in the grading of motor oil (in the US anyway). Clearly, this is the TLA of some other entity. Anyone know?

Also, the SAE Phonology Test is mentioned without any link. Is this an internet entity that may be viewed?

Thanks in advance!

[Ser]

Uhhh standard average European.

The phonology test is viewtopic.php?f=4&t=38945 I believe.

[garysk]

Thank you! I figured that it had something to do with European something, and now I recall reading the full phrase in other posts. And thanks for the link to the topic.

[k1234567890y]

I think there is actually at least one European grammatical feature that is not listed by them: the tendency to use "adposition+verb" compound verbs or phrasal verbs, Germanic languages, Slavic languages and Latin do all show such a way to create new verbs, although I am not sure the situations about Romance languages. In other languages, at least in Asian languages, "verb+verb" compound verbs seem to be more common.

I personally think the tendency to use "adposition+verb" compound verbs or "spatial adverb+verb" compound verbs can be listed as a SAE grammatical feature

[HazelFiver]

I think there is actually at least one European grammatical feature that is not listed by them: the tendency to use "adposition+verb" compound verbs or phrasal verbs, Germanic languages, Slavic languages and Latin do all show such a way to create new verbs, although I am not sure the situations about Romance languages. In other languages, at least in Asian languages, "verb+verb" compound verbs seem to be more common.

I personally think the tendency to use "adposition+verb" compound verbs or "spatial adverb+verb" compound verbs can be listed as a SAE grammatical feature

Celtic languages also have the adpositional verb prefixes (or had -- I'm not sure how productive they are anymore, and some of them no longer correspond to actual prepositions e.g. "di" and "cyf"), and I would guess that they are found in other IE branches as well. Celtic isn't considered part of SAE, though, so even that is enough to suggest that if anything it's an IE feature rather than SAE.

[mèþru]

I'm confused by 54. After careful reading of the WALS link in the document (23A), I thought I finally understood it, but now I'm not sure again. If a verb agrees with the subject in number and person while case on nouns are not marked/marked with prepositions depending on the case (completely unmarked on the verb), is it dependent marking, double marking or both? English, French and Spanish all have this situation and are shown in WALS as being dependent marking, no marking and double marking, respectively, without any examples.

[cromulant]

I'm confused by 54. After careful reading of the WALS link in the document (23A), I thought I finally understood it, but now I'm not sure again. If a verb agrees with the subject in number and person while case on nouns are not marked/marked with prepositions depending on the case (completely unmarked on the verb), is it dependent marking, double marking or both? English, French and Spanish all have this situation and are shown in WALS as being dependent marking, no marking and double marking, respectively, without any examples.

WALS 23A is strictly about how the Patient is marked. Subject marking is irrelevant. So if your language ever marks objects with a preposition, it is dependent marking. (Cases and adpositions are treated equivalently in this chapter, unlike other WALS chapters.) If it ever marks the object on the verb, it is head marking. If it does both of these things, it's double marking. Neither = no marking, even if it has an otherwise rich case system that doesn't distinguish A and P, and marks the A on the verb.

English should be "no marking." WALS isn't always consistent (or accurate).

[mèþru]

English marks the dative with prepositions. So do French and Spanish. As far as I'm aware, none of them mark verbs for this, so they shouldn't they all be dependent marking? If you mean by patient the O argument, shouldn't all of them be no marking?

[cromulant]

The chapter looks solely at direct objects. Dative are not part of the equation:

By removing the points most prone to follow universal tendencies, we can distill the morphosyntactic range down to two phrase types: possessive phrase with noun possessor (e.g. neighbor's house, the color of grass) and direct or primary object in the transitive clause (e.g. wrote books, broke a glass).

I don't know French. The Spanish preposition a used with human direct objects meets the "dependent marking" criterion, and don't Spanish object pronouns sometimes glom onto verbs? This may earn it "head marking" points as well.

But like I said. WALS is a great resource for learning about linguistic typology, and getting a broad strokes picture of the geographic distribution of certain features, but when you get down to specifics...I have seen a lot if data points on WALS that didn't check out, or that were logically incompatible with other data points (if you apply their criteria rigorously).

[mèþru]

I have an equivalent posposition to Spanish a in Khásoitoi/Proto-Kkasetean with mostly the same usage (including that specific one), but the object does not "glom" onto verbs (at least I think it doesn't... What do you mean by "glom"?) .What does that make my conlang?

[cromulant]

By "glom" I mean they are enclitics. This is the only way I can think of the Spanish could be said to mark objects on the verb--though they aren't used with finite verbs.

I don't know why WALS considers Spanish to meet the conditions (as defined by WALS) for head marking. It could be a mistake, or there could be criteria they are applying that they don't mention in the chapter.

Hard to say about your conlang without examples. It sounds like it is at least dependent marking, and probably not double marking.

[mèþru]

Thanks. I've figured it out.