[Lingtyp] Lexical nominalisation of property concepts

Thu Jun 23 08:27:13 UTC 2016

Hi Luigi,
I don’t think the answers are simple in the sense that language A does this and language B does that. Lots of things are possible in each language, and different adjectives or verbs might be treated differently. (For example, we can say in English That is my drink, but we don’t say That is my eat—nothing structural about that.) Markedness is for me not a structural notion in this case but just frequency (of course not having a marker is structurally less marked than having one). In Chinese there is also the option to use the head noun sè ‘color’ when one wants to refer to the color itself, but as I showed in my earlier msg to David that isn’t always necessary.

In Rawang, unlike in many other Sino-Tibetan languages, there is more of a distinction between noun and verb/adjective, and there are many ways that verbs/adjectives can be nominalized (see LaPolla, Randy J. 2008. Nominalization in Rawang. Linguistics of the Tibeto-Burman Area 31.2:45-66.  http://randylapolla.net/papers/LaPolla_2008_Nominalization_in_Rawang.pdf ). In the case of the example you gave, the plural marker functions as a kind of nominalizer, at least functionally. Classifiers can also be used for this. The topic marker is not crucial in this example, though. As in Chinese, making something the topic of the construction forces a referential interpretation, regardless of its form.

All the best,
Randy
-----
Prof. Randy J. LaPolla, PhD FAHA （羅仁地）| Division of Linguistics and Multilingual Studies | Nanyang Technological University
HSS-03-45, 14 Nanyang Drive, Singapore 637332 | Tel: (65) 6592-1825 GMT+8h | Fax: (65) 6795-6525 | http://randylapolla.net/

On 17 Jun 2016, at 9:53 pm, Luigi Talamo <luigi.talamo at unibg.it<mailto:luigi.talamo at unibg.it>> wrote:

Dear all,
thanks again for all your answers. I am glad that my questions has triggered such an hot and interesting discussion.
Like David, I have been fascinating with parts-of-speech (un)markedness since my introduction to linguistic typology.

As for marking strategies, in my research on 'referential properties', I have simply applied what has been done for 'predicative properties', that is, I have assumed that in Italian something like 'il buono' is structurally more marked than something like 'il cane buono'  'the good dog' and less marked than 'il cane è buono' 'the dog is good'. Still, we have some examples in Italian where predicative properties can be constructed without verbal copula and something like 'quello buono' 'good one', probably falls under this construction.

I may be wrong, but I think that from a cross-linguistic perspective we still do not have a clear description on how property concepts function as referents; in David's map, adjectives can be syntactically nouns in some languages, but they can be functionally either, as far as I can see from Randy's explanation of Mandarin 'hong de', predicative or referential.

Are there any languages that explicitly mark this difference i.e., which they employ a dedicated marker for property concept (PC) with referential function and with argument (i.e. not quality/abstract) meaning? I think that in Italian the distributional difference between 'il rosso' and 'quello rosso', that is, between a construction involving a simple article vs. a construction with a complex article (I employ Ch. Lyons 1999 distinction) goes in this direction; the former construction codes a PC with referential function, the latter a property concept with predicative function, which may grammatically act as a noun.
Moreover, things are clearer in argument nominal constructions such as 'Non posso sopportare queste stupid-aggini stupid-NMLZ', 'I cannot stand these stupid arguments'. Finally, note that the non-argument counterpart of 'quello rosso' is 'l'essere rosso' 'the being red', in which the predicative adjective red is nominalized by the means of a nominalized copula construction.

With respect to marking strategy 'heaviness' e.g., 'the definition of Mandarin de as a word ', the issue is not clearer in languages marking 'substantivization' with bound morphemes. For instance, in Lezgian, the substantizer -da/-di is employed both with PC with referential function:

(Haspelmath 1993:112)
I dünja.da-l qʰsan-bur pis-bur.u-laj gzaf ja.
this world-SRESS good-NMLZ.PL bad-NMLZ.PL-SREL many COP
'In this world the good (people) are more numerous than the bad (people).'

and PC with predicative function grammatically acting as nouns:

(Haspelmath 1993:111)
Hixtin televizor k'an-zawa wa-z? Gweči-di, č'exi-di, rang.uni-n-di?
which television want-IMPF you-DAT little-NMLZ.SG big-NMLZ.SG color-GEN-NMLZ.SG
'What kind of TV set do you want? A little one, a big one, a color one?'

whereas inflected property words are employed without further marking in Rawang (but, what about the TOPIC marker?):

Rawang (LaPolla & Poa 2001:258)

tē-rì nō shvnvt-í wvp yàng-à
large-PL TOP gun-INS shoot TOPyrs-3PL.PST
'The big ones were shot by gun.'

Still, I do not know how to classify tē-rì. Perhaps languages lacking an adjectival category or showing `verby' adjectives do not display an exact distinction between PC with referential functional and substantivized predicative PC.

Thank you, best.

Luigi

2016-06-13 12:24 GMT+02:00 David Gil <gil at shh.mpg.de<mailto:gil at shh.mpg.de>>:
Randy,

Thanks for your comments.  Two points:

With regard to whether Mandarin de is a separate word or not, your criticism is well-taken; my only defense is that that is the way it is usually characterized, and that in a typological survey of this scope, there is no other way of doing things other than to rely on extant descriptions.  Except perhaps to sidestep the issue of wordhood altogether and simply collapse "affix" and "separate word" into a single type, which, I suspect, is what would do now if I were doing the chapter all over again.

But I really don't see your point when you write: "I still don’t see what lumping together language forms that aren’t similar into categories that make them look similar does for us."  Surely this is the only way for rational inquiry into language (or any other phenomenological domain) to proceed.  "Similar" and "not similar" aren't binary holistic choices, they only have meaning in the context of particular criteria or properties.  We observe two entities, call them A and B, and then say Hey, A and B are alike with respect to property X.  The value of saying this depends on how trivial or insightful the property X turns out to be, ie. what further understandings X leads us towards.  But crucially, the value of X is not negated by pointing to properties Y, Z, W, V etc, with respect to which A and B differ.  The existence of such properties with respect to which A and B differ is totally irrelevant to the value of property X, they do not impinge on it in any way.

You ask "what has lumping Mandarin and English together in this context taught us about the languages?".  Well one of the things I've always been interested in is cross-linguistic variation with respect to parts-of-speech inventories.  The present WALS map addresses the issue of whether a language distinguishes between adjectives and nouns.  (Note: I'm saying "addresses", not "answers".)  Specifically, if a language, like English or Mandarin, needs to add a grammatical marker to an adjective in order to give it the distributional properties of a noun, then this provides good reason to suspect that in such languages, adjectives and nouns constitute different word classes, defined distributionally.  Whereas if a language, like Italian or Hebrew, doesn't need to make use of such a marker, then perhaps it doesn't distinguish between adjectives and nouns (as indeed is suggested by the traditional term "substantives" that groups the two classes together), though alternatively it could be the case that the language in question does distinguish between adjectives and nouns using other criteria.

So all this is relevant to English and Mandarin, regardless of the myriad other important differences between English one and Mandarin de.

Best,

David

On 13/06/2016 17:44, Randy John LaPolla (Prof) wrote:
Hi David,
Thanks for your reply. The crux may be the definition of Mandarin de as a word (you don’t specify phonological word or grammatical word, but since you treat clitics—grammatical words that aren’t phonological words—differently, I am assuming you mean phonological word). It cannot appear on its own, and when added to another word, like hong, they are pronounced together, so it patterns like a clitic, and so is unlike English one in that way as well (people are often thrown off by the fact that in Chinese each character is written separately, but that doesn’t mean each character is a phonological word).

And although I don’t want to start the whole debate we had in January again, I still don’t see what lumping together language forms that aren’t similar into categories that make them look similar does for us. Although I can see the practical difficulties of taking the actual facts of all the languages seriously, very concretely, what has lumping Mandarin and English together in this context taught us about the languages?

Thanks very much.

All the best,
Randy

On 12 Jun 2016, at 1:36 pm, David Gil <<mailto:gil at shh.mpg.de>gil at shh.mpg.de<mailto:gil at shh.mpg.de>> wrote:

Randy,

Yes, my chapter in WALS characterizes the English and Mandarin constructions as "of the same type structurally", and yes, the two constructions are different from each other in precisely the ways that you describe!

That's what typology does: dividing things into classes according to one set of criteria, thereby putting in to the same class things that are very different according to other sets of criteria.  And that's precisely what has happened here.  My WALS chapter asks whether an adjective can occur on its own as a noun, without any further morphosyntactic marking and the answer for both English and Mandarin is the same: no.  It then further asks, for languages that require such morphosyntactic marking, what the formal properties of the marking is, distinguishing between affixes and separate words, and between forms that occur before and after their host adjective.  And once again, Mandarin and English come out the same, with a separate word that occurs after its host adjective.  That's all the WALS chapter purports to say.

Now clearly many constructions in different languages with the same WALS feature values will differ from each other in myriad other ways, as is the case for English and Mandarin here.  You may feel that the typology proposed in the "Adjectives without Nouns" WALS map overlooks what's "most important" about the constructions in question, and you could indeed be right about that.  I suspect, however, that an alternative "Adjective without Nouns" map distinguishing between "English and Mandarin types" on the basis of headedness would have been impractical to produce, since it is too theory dependent, and hence it would not have been possible to glean the necessary information from available grammatical descriptions of a sufficiently large sample of languages.  (In fact, while I agree entirely with your description of the difference between English and Mandarin, I bet that there are even grammatical descriptions of English and Mandarin out there that would see things differently.)

I hope this clarifies matters ...

David

On 12/06/2016 08:20, Randy John LaPolla (Prof) wrote:
Hi David,
It seems from your message here and from your chapter in WALS that the English construction with one and the Chinese construction with de are of the same type structurally. I don’t know if I have read you right, but although they are made up of the word representing a property concept followed by another word, the two constructions are quite different (and the natures of all of the words involved are different as well). In the relevant use of English one, it is a pro-form (see  Goldberg, Adele E. & Laura A. Michaelis. 2016. One among many: anaphoric one and its relationship to numeral one. Cognitive Science 40.4:1–26. DOI: 10.1111/cogs.12339  for interesting discussion) and clearly the head of the phrase, but in the Chinese example de is only a nominalizer and clearly not the head of the phrase, either in terms of structural behaviour (e.g. in English one patterns like other heads, e.g. we can say “this one”, but this is not the case with Chinese de) or in terms of speakers’ “feel” for what is the core element of the phrase.

This sort of goes back to the discussion on categorization we had back in January.

All the best,
Randy
-----
Prof. Randy J. LaPolla, PhD FAHA （羅仁地）| Division of Linguistics and Multilingual Studies | Nanyang Technological University
HSS-03-45, 14 Nanyang Drive, Singapore 637332 | Tel: (65) 6592-1825 GMT+8h | Fax: (65) 6795-6525 | <http://randylapolla.net/> http://randylapolla.net/

On 11 Jun 2016, at 3:33 pm, David Gil <gil at shh.mpg.de<mailto:gil at shh.mpg.de>> wrote:

Luigi,

Unlike many of my typologist colleagues who seek refuge from the muddy waters of formal criteria in the supposed clarity of semantics, I find semantic criteria to often be just as problematical, if not more so, than their formal counterparts.

For the purposes of my WALS map, I did not use headedness as a defining criteria, and I would not wish to take a stand on the headedness in the examples that you discuss.  By "adjective" I meant property-denoting word one of whose typical functions is as an attribute of a noun, and by "noun" I meant thing-denoting word.  The map shows the morphosyntactic strategies that a language uses to allow an adjective to occur in a noun slot — typically, but not criterially, heading a phrase that occurs in an argument position.  This definition is met, among others, by the one in English beautiful one, the de in Mandarin hong de, and also by the lack of (dedicated adjective-to-noun conversion) marking in the Italian il bello.

Best,

David

On 10/06/2016 23:01, Luigi Talamo wrote:
Dear all,
thanks a lot for your all answers, I really appreciate that.
I have found your data very interesting, many comments will follow :-)
I begin below with David's answer.

One of the two kinds of nominalization mentioned in the query ('beautiful' > 'beautiful one') is the subject of my WALS map #61 "Adjectives without Nouns".

David

Thanks David, I have read your WALS map at the beginning of my work; maybe you remember that we have exchanged a couple of e-mails some time ago. As you mention in the WALS article, the most important issue here is whether adjectives are syntactic heads in constructions such as 'the white one', which translates in Italian as 'quello bianco'. As you probably noticed, I did not consider these constructions in my study, as they appear to me to be more 'predicative' than 'referential', at least in Italian; moreover, the syntactic head of the Italian construction is most likely the deictic quello 'this'. But what about the Mandarin example that is reported in your map, Wǒ yào hóng de. ? Is hóng a property concept with referential function ?

Thanks

Luigi

On 09/06/2016 21:14, Luigi Talamo wrote:
Dear all,
I am conducting a research on the lexical nominalisation of property concepts in contemporary Italian. My study involves two types of nominalisation strategy, affixation such as bello `beautiful' -> bell-ezza `beauty (abstract concept)' and zero-marking ('conversion'), such as bello (adj) -> `(il) bello' -> `the beautiful person', `beauty (abstract concept)' and `what is beautiful about something'.
Drawing mostly from 'Leipzig Questionnaire On Nominalisation and mixed Categories' (Malchukov et alii (2008)) and studies on adjectival and mixed categories, I have elaborated a series of morpho-syntactic and semantic parameters, which I have employed to study de-adjectival nominalizations in actual, corpus-based contexts.
I would like to insert in my study some cross-linguistic notes on the phenomenon, which I hope to further study from a typological perspective. I will be glad if you can provide me some examples from your languages of expertise. I have found some examples of de-adjectival nominalizations here and there in grammars, but I was not able to exactly figure out which are the parameters involved; moreover, some recent works (among others, Roy (2010), Alexiadou et alii (2010), Alexiadou & Iordachioaia (2014)) give interesting insights on de-adjectival nominalization, but examples are limited to European languages.

I am particularly interested in non-European languages showing a distinct class of adjectives; morpho-syntatic parameters include case, number, gender, definiteness and specificity, degree, external argument structure and, possibly, verbal parameters, which are however not very significant for Italian de-adjectival nominalisation; semantic parameters include referent animacy, the distinction between the nominalisation of the adjectival 'argument' vs. the nominalisation of the adjective itself e.g., softie `a thing which is soft' vs. softness and the semantic type of property concepts e.g., PHYSICAL PROPERTY or HUMAN PROPENSITY.

So, possible questions are as following:
1. Can property concepts be turned into nouns?
2. Which strategies are employed for this purpose?
3. Which parameters do de-adjectival nouns display?
4. Are there any missing values for a given parameter? For instance, de-adjectival nouns can be only singular or definite or restricted to the subject position.
5. Are de-adjectival nouns found in both semantic types of nominalization? For instance, I have observed that European languages focus on the nominalisation of the adjective itself, while argument nominalizations are scarcely attested, limited to certain language varieties and not stable in the lexicon.

(needless to say, questions 2 to 4 can have multiple answers, helping to describe different patterns of property nominalisation)

Thanks in advance for your help, all the best.

Luigi

--
PhD Program in Linguistics ('Scienze Linguistiche')
University of Bergamo and University of Pavia - Italy

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

--
David Gil

Department of Linguistic and Cultural Evolution
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10, 07745 Jena, Germany

Email: gil at shh.mpg.de<mailto:gil at shh.mpg.de>
Office Phone (Germany): +49-3641686834<tel:%2B49-3641686834>
Mobile Phone (Indonesia): +62-82238009215<tel:%2B62-82238009215>

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

--
PhD Program in Linguistics ('Scienze Linguistiche')
University of Bergamo and University of Pavia - Italy

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

--
David Gil

Department of Linguistic and Cultural Evolution
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10, 07745 Jena, Germany

Email: gil at shh.mpg.de<mailto:gil at shh.mpg.de>
Office Phone (Germany): +49-3641686834<tel:%2B49-3641686834>
Mobile Phone (Indonesia): +62-82238009215<tel:%2B62-82238009215>

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

________________________________
CONFIDENTIALITY: This email is intended solely for the person(s) named and may be confidential and/or privileged. If you are not the intended recipient, please delete it, notify us and do not copy, use, or disclose its contents.
Towards a sustainable earth: Print only when necessary. Thank you.

--
David Gil

Department of Linguistic and Cultural Evolution
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10, 07745 Jena, Germany

Email: gil at shh.mpg.de<mailto:gil at shh.mpg.de>
Office Phone (Germany): +49-3641686834<tel:%2B49-3641686834>
Mobile Phone (Indonesia): +62-82238009215<tel:%2B62-82238009215>

--
David Gil

Department of Linguistic and Cultural Evolution
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10, 07745 Jena, Germany

Email: gil at shh.mpg.de<mailto:gil at shh.mpg.de>
Office Phone (Germany): +49-3641686834<tel:%2B49-3641686834>
Mobile Phone (Indonesia): +62-82238009215<tel:%2B62-82238009215>

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

--
PhD Program in Linguistics ('Scienze Linguistiche')
University of Bergamo and University of Pavia - Italy
_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20160623/b54373d7/attachment.htm>