SAE phonology and grammar tests

[Salmoneus]

Well, phthisie is an old spelling. There's also phtalate (English phthalate). Yes, very rare, but the French man up and plow through those initial clusters (cf. pneu, psychologie, ptérodactyle) where we skip the first consonant.

Speak for yourself! I've always heard phthisis and phthalate with /fT/ clusters. And sometimes I hear the /p/ in /ps/ and /pt/ clusters, where the word is specialised - don't think I've ever heard a /pn/ initial cluster, though. And <chth> I've only heard as /kT/, with the first fricative being stopped, presumably because it's one we don't have.

[Curlyjimsam]

Viksen (at least "Viksen 2010 edition", which is the most recent "complete" grammar ... maybe it's time for an official updating) scores about 21 on grammar, though I was unsure on some of the points (because I never explicitly addressed them or because I was unclear whether something would fit the criteria or not).

As I would have predicted, however, it scores much higher on the phonology one: 61 (I never was able to bring myself to get rid of all those fricatives ...).

EDIT: Now also tried with the "revised phonology" I toyed with about 5 years ago, where I did get rid of the fricatives (amongst other alterations), but which I never got round to actually making "official". That scores 38.

[HazelFiver]

Viksen (at least "Viksen 2010 edition", which is the most recent "complete" grammar ... maybe it's time for an official updating) scores about 21 on grammar, though I was unsure on some of the points (because I never explicitly addressed them or because I was unclear whether something would fit the criteria or not).

As I would have predicted, however, it scores much higher on the phonology one: 61 (I never was able to bring myself to get rid of all those fricatives ...).

EDIT: Now also tried with the "revised phonology" I toyed with about 5 years ago, where I did get rid of the fricatives (amongst other alterations), but which I never got round to actually making "official". That scores 38.

38 is extremely low. 61 is already low enough to be decidedly un-European. From the original thread, Japanese, Mandarin and Ojibwe all scored higher than that. Unlike the grammar test, the phonology test seems to be measuring where a language is on the spectrum from SAE to its opposite. Speaking of fricatives, the phonology I posted above has 32 of them, FWIW.

[HazelFiver]

40. CCV syllables allowed but SSV or SSCV syllables not allowed (not including geminates)

Slavic languages are considered peripherally SAE in Haspelmath's paper, and they allow SSV. However, they require the second plosive to be coronal, so they have /pt bd kt gd/ but not */pk bg/. The IE preference for coronals was noted by someone on the phonology test thread, and this is a manifestation of that. I suggest changing this to:

40. CCV syllables allowed but SSV or SSCV syllables not allowed (not including geminates) [half mark if the second S must be coronal]

Scratch that. Apparently Russian allows initial /tk/ (example). I still don't think they allow the clusters I actually named though. If anything this should be "if one S must be coronal". I should really do some research before I start talking.

I'd still like to hear what you think of my comments on questions other than the fricative cluster ones. I could just go ahead and change stuff around, but it would be nice to get some kind of feedback or consensus first.

[Pabappa]

Sorry but I think you may need to modify the criteria yet again, because Russian also has a preposition, https://en.wiktionary.org/wiki/%D0%BA#Russian , which is pronounced as just a /k/, binds to the following syllable, and can occur before plosives such as /p/ and /b/, therefore allowing initial /kp/ in the onset. The variant form with the epenthetic /o/ appears only before certain clusters.

[Frislander]

I wouldn't change test questions on the basis of Russian, because everyone agrees that Russian is at best marginally SAE.

[Zju]

40. CCV syllables allowed but SSV or SSCV syllables not allowed (not including geminates)

Slavic languages are considered peripherally SAE in Haspelmath's paper, and they allow SSV. However, they require the second plosive to be coronal, so they have /pt bd kt gd/ but not */pk bg/. The IE preference for coronals was noted by someone on the phonology test thread, and this is a manifestation of that. I suggest changing this to:

40. CCV syllables allowed but SSV or SSCV syllables not allowed (not including geminates) [half mark if the second S must be coronal]

Scratch that. Apparently Russian allows initial /tk/ (example). I still don't think they allow the clusters I actually named though. If anything this should be "if one S must be coronal". I should really do some research before I start talking.

I'd still like to hear what you think of my comments on questions other than the fricative cluster ones. I could just go ahead and change stuff around, but it would be nice to get some kind of feedback or consensus first.

Not having (maybe? can't think of counter examples at least) /pk bg/ is just chance though, with a proper placement of yers in PSl you can get a surprising variety of clusters. PSl. *pъkъ or *bъgъ would have yielded /pk/ or /bg/ respectively in at least some Slavic languages. Also consider that the word for bee contains either /pt͡ʃ/ or /bd͡ʒ/ in some languages of the family and that postalveolars invariably come from velars.

[Boşkoventi]

Polish has several words with /kp/ and /gb/, all apparently related to/derived from these:
kpić [kpitɕ] "jeer, sneer, mock"
gbur [ɡbur] "rude, boor, yokel"

And there are plenty of other fun clusters:
ptak [ptak] "bird"
kto [kto] "who"
dbać [dbatɕ] "care, take care, look after"
gdy [ɡdɨ] "when"
tknąć [tknɔntɕ] "touch, strike, seize"
lśnić [lɕnitɕ] "glitter, shine"
lniany [lɲanɨ] "linen, flax"
lgnąć [lɡnɔntɕ] "adhere, sink, stick"
łbica [wbitsa] "cylinder head"
łkać [wkatɕ] "sob, weep"
łza [wza] "tear" (n.)

Factor in all the fricatives and affricates, and I find the claim that it's even peripherally SAE more than a little dubious.

[HazelFiver]

Okay, looks like we can forget about that one. However, my point about the syllabic rhotic question still stands. If it's not there to make certain Slavic languages score higher, what's it for?

Factor in all the fricatives and affricates, and I find the claim that it's even peripherally SAE more than a little dubious.

More fricatives and affricates make it more SAE, not less. As for the consonant clusters, there are other factors that contribute to an SAE-like phonology. The test also doesn't give much relative weight to non-SAE clusters -- I mentioned earlier on this thread that I constructed a phonology that has unlimited consonant clusters and it scores 86. Not that I know much about Polish.

[Nortaneous]

Well, phthisie is an old spelling. There's also phtalate (English phthalate). Yes, very rare, but the French man up and plow through those initial clusters (cf. pneu, psychologie, ptérodactyle) where we skip the first consonant.

Speak for yourself! I've always heard phthisis and phthalate with /fT/ clusters. And sometimes I hear the /p/ in /ps/ and /pt/ clusters, where the word is specialised - don't think I've ever heard a /pn/ initial cluster, though. And <chth> I've only heard as /kT/, with the first fricative being stopped, presumably because it's one we don't have.

I usually hear <c(h)th> with an epenthetic schwa, but people are usually weak

[Xwtek]

What if ejective occurs as allophone

[Xwtek]

Also the only European language with word final -h is Finnic. (Maybe Saamic?)

[anteallach]

Also the only European language with word final -h is Finnic. (Maybe Saamic?)

Both Irish and Scottish Gaelic (not generally regarded as SAE, I know) allow word final /h/, written <th>.

Also there are cases of final [h] developing elsewhere in Europe, including in some dialects of Spanish (allophone of /s/) and Liverpool English (arguably an allophone of /t/).

[Xwtek]

Also the only European language with word final -h is Finnic. (Maybe Saamic?)

Both Irish and Scottish Gaelic (not generally regarded as SAE, I know) allow word final /h/, written <th>.

Also there are cases of final [h] developing elsewhere in Europe, including in some dialects of Spanish (allophone of /s/) and Liverpool English (arguably an allophone of /t/).

Well, the point is it's uncommon. However, the test is saying as /h/ is allowed word finally.

[Nortaneous]

What if ejective occurs as allophone

I think this is limited to English (and maybe Danish?) dialects.

[Xwtek]

What if ejective occurs as allophone

I think this is limited to English (and maybe Danish?) dialects.

I mean, does it count as violating?

[Imralu]

I did the test for Ngolu / Iliaqu and found it much easier than several years ago, thanks to the better wording and examples.

The results:

Grammar 20%
Phonology 46%

In the grammar, the only ones I ticked were:

• Differentiation between intensifiers and reflexive pronouns
• No productive use of reduplication
• Obligatory definite and indefinite articles/suffixes
• Conjunction "A and-B" i.e. "and" does not precede or follow the whole list and is not a suffix
• Topic and focus expressed by intonation and word order, not particles and affixes.
• Nominative-accusative morphosyntactic alignment
• Only one converb (non-finite subordinate verb), preference for finite subordinate clauses

I find the wording of the "converb" question a bit confusing because whenever I see "verb" I think "lexical verb" and not "verb form", so for a while I was trying to think of which verb gets used as a converb in SAE languages ... "Is it be? That's common! Oh ... converb form? That would make more sense."

Also, Ngolu / Iliaqu has two different words for "and", one used to link clauses (xe) and one used within a verbal phrase (xue). No word for "and" is present when conjoining noun phrases - noun phrases are simply juxtaposed in the same case, e.g.

uoia

sing

xue

and(VP)

bata

dance

nu

NOM.1S.ICS

I sing and dance.

uoia

sing

nu

NOM.1S.ICS

ja

NOM.3S,DEF.ACS

-s

-E

oko

dog

The dog [and] I sing.

mala

house

uju

GEN.3S.DEF.ICS

lima

married.woman

uju

GEN.3S.DEF.ICS

muja

initiated.man

The woman [and] the man have a house.

So, yes, when there is a word for "and" present, it goes between the words it conjoins (in fact, between all of them, A and B and C), but I wouldn't exactly call it SAE. Using only identical cases to coordinate noun or pronoun phrases is not something I can imagine in an SAE lang.

What in German would be expressed as: Dieser Mann und ich haben dich gesehen. (= That man and I saw you) would be, to use the Ngolu rule in German, would turn out like: *Dieser Mann haben ich dich gesehen. (I put haben in between, since the two nominative noun phrases need not be next to each other in Ngolu, so in German they would also not function as one unit and German's V2 word order would end up separating them in this instance.)

What I think is SAE is having a distinct separate word for "and" which does not also mean "with" in the comitative sense and is used to join elements at all kinds of levels within the sentence (whole clauses, noun phrases, verb phrases, prepositional phrases, adjective phrases). I'm not looking at WALS right now, but non-SAE languages often conflate "with" and "and", and/or have different rules for different sentence levels. E.g. in Māori, adjectives are conjoined only by repeating the noun or using a dummy noun like "thing" ... you cannot say "a big and old house" only "a house big, a house old" or "a house big, a thing old". I mean, check WALS and other things before changing it, but that's what I'm thinking at the moment.

I usually hear <c(h)th> with an epenthetic schwa, but people are usually weak

I heard my also-Australian but non-Germanophone flatmate recently threw the word Kneipe into an English sentence, pronounced with a schwa to break up the /kn/ and for a moment I didn't understand the word. I thought of canapé first.

The only word I knew before reading this thread with an initial /kθ/ is Cthulhu and don't need epenthesis to get through that.

What if ejective occurs as allophone

I think this is limited to English (and maybe Danish?) dialects.

I mean, does it count as violating?

I take all of the questions to be asking about phonemic distinctions rather than allophonic distinctions unless otherwise specified. I think that was the original intention. Otherwise, English has a whole lot of less SAE stuff: voiced, voiceless unaspirated and voiceless aspirated plosive series, velar plosives (e.g. as an allophone before my universally dark /l/), coarticulated consonants ([ʔ͡t] in "football", [ɡ͡b] in "logbook"), ejectives, syllabic consonants [n̩ ɻ̩ ɫ̩] etc. etc. etc.

[Birdlang]

Slarg got a 25 in phonology and 22 in grammar.
I didn’t put one of them in because I knew it would get a 0 in both. It has consonants like /ɢˁ͝ʒʶʷ/ in it.

[Zju]

My unnamed conlang scored about 66% on phonology (more than I'd expect) and 39% on grammar. (little less than my guesstimate)

What is a converb?

9. Ignoring stops and nasals: more fricatives than non-fricative consonants [2.5 marks if equal]

Are fricatives ignored or not?

[HazelFiver]

It's been a while. Bumping this thread because it was replied to after I vanished into thin air.

My unnamed conlang scored about 66% on phonology (more than I'd expect) and 39% on grammar. (little less than my guesstimate)

What is a converb?

Maybe this still isn't explained well enough.

a nonfinite verb form that serves to express adverbial subordination: notions like 'when', 'because', 'after' and 'while'.

Examples:

*On being elected president, he moved with his family to the capital.
*He walks the streets eating cakes.

English only has the -ing form, which also serves as a gerund. Sparrowgrass has multiple converbs (at least, I think that's what they are), e.g.
*Sgovižumit šov si atosona. I talked to him/her for a while as we waited.
*Rŷvošosit fjosb, ngukusit lamp. After opening the door, s/he entered the room.

This is what Haspelmath's paper says about it (mutilated by OCR):

European languages tend to have just
one converb (- Art. 83) (cf. Nedjalkov 1998).
For instance, Romance languages have the
gerundio I gt ronrlrf, English has the -lng-form,
and Slavic and Balkan languages have their
adverbial participle. The Celtic languages in
the west completely lack such a fonn, and the
languages east of SAE tend to have more
than one converb. Otherwise the core European languages tend to have adverbial conjunctions (' Art.63) to make adverbial
clauses. According to Kortmann (1997: 344\,
they have "a large, semantically highly differentiated inventory of free adverbial subordinators placed in clause-initial position".
More generally, they tend to have finite rather
than non-finite subordinate strategies (r
Art. 100), though a multi-purpose infinitive
usually exists (except lbl the Balkan languages

(By the way, the paper seems to have gone offline. I retrieved it from the Wayback Machine.)

9. Ignoring stops and nasals: more fricatives than non-fricative consonants [2.5 marks if equal]

Are fricatives ignored or not?

You wouldn't ignore fricatives for a question that asks the number of fricatives.