7

As far back as the mid 1700s, William Ward considered the following phrase in An Essay on Grammar applied to the English Language.

  • the flowers which a lady sitting on the seat in a garden views with attention

Ward noted that the example clearly shows how words like which “remove from the place of the word of the original sentence in order to stand as close as possible to that word considered as the antecedent to the relative when the sentence is turned into a relative clause depending on such antecedent.” Whether he conceives of this as a procedure that we undertake or simply as a useful description is unclear. Either way, he speaks about movement.

Was this kind of explanation used in other grammars, perhaps as far back as Panini, or early Greek or Latin grammars? What's the earliest example of this kind of thinking?

  • The Greeks no doubt knew the simple fact that a relative normally needed to be close to its antecedent, sometimes overriding other rules of syntax. But you're probably looking for something more specific? – Cerberus Jul 22 '16 at 18:12
  • Yes, I expect you're right, but I'd like something more specific. – Brett Reynolds Jul 22 '16 at 18:23
  • I think the concept is as old as Greek Antiquity; people know how to use their language and teach (the title of your question). As to what the earliest Greek or Mesopotamian source is that explicitly describes it, I don't know (the end of your question); let's hope someone finds a good quotation. – Cerberus Jul 22 '16 at 18:28
  • That is a great quote from William Ward, but I wonder it it's right. In the view of the relative clause construction taken in Generalized Phrase Structure Grammar, the reason that the phrase with the relative pronoun turns up at the beginning of the relative clause is because it is topicalized. It has nothing to do with the relative pronoun being next to the word that the relative clause modifies. And anyway, it's not always next to it. – Greg Lee Jan 13 '17 at 20:19
  • I'm not claiming it's right. I'm just curious about when people started thinking about it in this way. – Brett Reynolds Jan 14 '17 at 18:32
4

What is different between English on the one hand, and such ancient Indo-European languages as Latin, Greek and Sanskrit, is that the latter had a very free constituent order: not only the order of constituents within the sentence was free, but even the word order within a phrase was almost unrestrained. (Incidentally, this why reading Latin poetry can be often so tough.)

With a similar linguistic background it is clear why Ancient grammarians did not develop any notion of syntactic structure that could be comparable with the one developed by modern syntacticians who initially used mainly English as their only object-language. I personally would say that Greek and Sanskrit lacked syntactic constituents at all (but I imagine there are many who would object). What these languages obviously did not lack was the semantic dependencies between words (such as the dependency of the arguments from the verb). Indeed, Pāṇini developed a very refined theory of what the Generativism calls "theta-roles" and Charles Fillmore used to call "deep cases". On the other hand, there is no mention of syntactic constituency in Pāṇini at all, to my knowledge. Not to mention Greek and Latin grammarians who usually were not interested in anything else than morphology and semantics (and a bit of phonology).

UPD: This book by Graffi could contain pertinent information concerning the period from the beginning of the XIX century onwards.

| improve this answer | |
  • 1
    So are you saying, pertinent to the original question, that the ancients had no notion of the movement of words out of their original constituents, because they hadn't figured out that there were constituents there to move out of? (Sort of like modern dependency grammarians.) – Greg Lee Jan 13 '17 at 19:18
  • 1
    Sorry, probably I was not very explicit in distinguishing these two aspects: 1) to my knowledge, no ancient grammarian ever speaks about syntactic movements; 2) this is so because the language whose grammar they were describing was a free word order type of language with apparently no constituency and no movement. I am quite certain concerning the first point and the second point appears to me as a good explanation of this fact. – Artemij Keidan Jan 13 '17 at 19:34
  • All the anciently grammatized languages (Latin, Greek, Sanskrit) were highly inflected languages whose words were coded for agreement, making word order available for poetry and emphasis, the way it is in Slavic languages (also highly inflected). One situation pretty much goes with the other. – jlawler Jan 14 '17 at 3:49
2

It seems that the Mashi Wentong (1898) discusses fronting. "The “normal” subject is always called qici , while the topic is called zhuci, as can be seen in the analysis of sentence (4), where Zhuanyu is labeled zhuci, while later, in the commentary on shi 是, it is referred to as a “fronted qici”" (p. 66). Obviously, this is later than Ward, but perhaps worth noting.

Peverelli, P. (2015). The history of Modern Chinese grammar studies. Berlin, Heidelberg: Springer-Verlag. http://doi.org/10.1007/978-3-662-46504

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.