1101. IrvingSnodgrass - Jan. 24, 1999 - 6:30 AM PT
I'd like to continue with my look at the Austric Languages, if anyone is interested. (If not, there's a handy scroll bar).
As you may recall, the Austric group, as depicted in the chart (linked to above), contains four language families. Yesterday I breezed through three of these families (Miao-Yao, Austro-Asiatic, and Daic). Now I'd like to take a look at my favorite language family, Austronesian.
The Austronesian family has more languages (959) than any other language family, and includes languages on virtually every island in the Indian and Pacific Oceans and in between, as well as a few outlier languages on the mainland. The relationships among the languages are of particularly interesting character, as I'll try to show.
The Austronesian family has four sub-families. Three of these sub-families (Atayalic, Tsouic, and Paiwanic), accounting for 14 languages, are found on the island of Taiwan, and represent the aboriginal (pre-Chinese) inhabitants of the island. The island of Taiwan is also the likely homeland of all Austronesians, by virtue of it being the location of greatest linguistic diversity. (The chart has a small error, which I'm too lazy to correct: there should be *4* Tsouic languages, not three). The other sub-family, containing 945 languages spoken by around 400 million people, is the Malayo-Polynesian sub-family.
One interesting fact about the Malayo-Polynesian sub-family is that the relationships between the 945 languages are transparent, and can be seen even by the most ignorant non-linguist. The word for "five," for example, in Indonesian, Tagalog and Hawaiian (the easternmost of all the languages) is identical: "lima." In fact, nearly all the languages in this far-flung group have cognate terms for this one word (usually "lima" or "rima").
[continued]
1102. IrvingSnodgrass - Jan. 24, 1999 - 6:40 AM PT
I once came across a pamphlet prepared by a linguist at the University of Hawaii which was a handy guide to converting Indonesian words to their Hawaiian equivalents. With a few simple sound changes, many words are predictable (Indonesian "langit" --> Hawaiian "lani," meaning "sky"). I was able to teach myself the Cebuano language of the Central Philippines using the same method (Indonesian "r" becomes Cebuano "g," Indonesian "d" becomes Cebuano "l": tidur --> tolog "sleep", beras --> bugas "rice").
Let's have a look at the Malayo-Polynesian sub-family.
As you can see, there are only two branches of Malayo-Polynesian: a Western branch with 374 languages, and a Central-Eastern branch with 571 languages.
There are 8 sub-branches from the Western branch, with four branches based in the Philippines, and four branches based in Indonesia. Of the 135 languages in the four Philippino sub-branches, only three are not found in the Philippines (the Chamorro, Palauan and Yapese languages of Micronesia). On the other hand, the 236 languages of the Indonesian sub-branch are found in 9 nations.
The two best known languages of the Philippines branches are Tagalog, the national language of the Philippines, and Cebuano the language with the most native speakers in the Philippines. Another language, Chamorro, is the official language of Guam, while Palauan is the official language of the Republic of Palau.
[continued]
1103. IrvingSnodgrass - Jan. 24, 1999 - 6:46 AM PT
In the Indonesian sub-branches, the "Celebes" sub-branch includes 61 languages found on the island of Celebes (Sulawesi), with a few in the Southern Philippines as well. The best-known of these languages are Makassarese and Bugis, the languages of sea-faring groups who have had significant impact in SE Asian history, and Gorontalo, the native language of Indonesia's current president (for a few more months) BJ Habibie.
A smaller branch is the Sama-Bajaw branch, with 20 languages spoken in the Southern Philippines, and both the Malaysian and Indonesian parts of Borneo.
The remaining two Indonesian branches are where things start to get interesting. The Borneo branch includes 109 languages, *most* of which are spoken on the island of Borneo (both the Indonesian and the Malaysian parts).
However, there is one language which is spoken in a completely different nation. In fact, this language has 8 million speakers and is the national language of this nation. Let's see who can find it first (it's on this chart). I'll provide historical details later.
Also later, I'll take a look at the Sundic sub-branch, the largest by far in terms of native speakers and influence of any Austronesian group. I'll have a few interesting facts to point out there as well.
1104. DanDillon - Jan. 24, 1999 - 7:42 AM PT
One of your favorite topics, indeed! Please go on.
For some reason unknown to me, I can't access your jpg charts. A new browser window pops up, but the screen comes up blank. But I'm enjoying your vast store of knowledge nonetheless.
1105. IrvingSnodgrass - Jan. 24, 1999 - 8:18 AM PT
Dan:
The jpegs should be working fine. They work on my screen, anyway.How can you answer the question in Message #1103 if you can't see the chart? Is anyone else having trouble?
1106. arkymalarky - Jan. 24, 1999 - 8:45 AM PT
Works fine for me.
1107. IrvingSnodgrass - Jan. 24, 1999 - 9:09 AM PT
Cool, Arky. So, did you find the answer?
1108. arkymalarky - Jan. 24, 1999 - 9:22 AM PT
Ha. No, I was just skimming before breakfast and when I saw Dan's remark I checked the link to see if it worked. I'll look now.
1109. pseudoerasmus - Jan. 24, 1999 - 9:25 AM PT
Something in Madagascar
1110. arkymalarky - Jan. 24, 1999 - 9:26 AM PT
The only language I recognize is Tagalog. I'm out of my element here.
1111. pseudoerasmus - Jan. 24, 1999 - 9:31 AM PT
It has to be Malagasy of Madascar, off the coast of Africa in the Indian Ocean.
Of course, if it's true, then that would be really really fascinating.
1112. IrvingSnodgrass - Jan. 24, 1999 - 9:31 AM PT
I see I linked the wrong chart when I asked my question. It should have been this chart of the languages of Borneo. All but one of the languages on this chart are found in Borneo. The other language is a surprise.
1113. pseudoerasmus - Jan. 24, 1999 - 9:34 AM PT
Madagascar.....
If Malagasy is a Malayo-Polynesian language, is it a remnant from the time Africa and Australasia were conjoined, or does the Malayo-Polynesian speaker population represent a remigration to Madagascar? You know, those wily seafarers, they went as far east as Hawaii and (allegedly) Easter Island. So why not go back west as well?
1114. IrvingSnodgrass - Jan. 24, 1999 - 9:35 AM PT
PE:
Correct. And it is true. The closest relative to Malagasy is the Ma'anyan language of the Eastern coast of Borneo. It seems that at some point in the past 2000 years, some seafarers from a small village in Borneo crossed the Indian Ocean and settled Madagascar.
1115. IrvingSnodgrass - Jan. 24, 1999 - 9:38 AM PT
PE:
Your second theory is the correct one. And it is a relatively recent migration, hence the placement in a sub-grouping of a sub-branch of a sub-family in the Austronesian family. Ma'anyan and Malagasy are as closely and transparently related as, say, Dutch and German.
1116. pseudoerasmus - Jan. 24, 1999 - 9:42 AM PT
It's hardly to believe that a bunch of villagers from Borneo travelled to Madagascar! This took place presumably thousands of years ago (to account for the linguistic evolution)?
This aspect of Madagascar intrigues me. Given the strong Arabic influence along the entire East Africa, and the Malayo-Polynesian base of the population, Madagascar must have an strong Asiatic "feel" to it, no? Add the old French colonial status, and it becomes an addition to my list of exotic locales to visit.
1117. IrvingSnodgrass - Jan. 24, 1999 - 9:50 AM PT
PE:
For a linguistic relationship this close, you certainly don't need thousands of years. Ma'anyan and Malagasy are nearly mutually intelligible. I would put the settlement of Madagascar closer to 1000 years ago, on a linguistic basis.
Here is the Ethnologue entry for Madagascar. As you can see, the classification is as in my chart: Malayo-Polynesian, Western
Malayo-Polynesian, Borneo, Barito, East, Malagasy. Madagascar hasn't even been settled long enough for the various dialects to evolve into separate languages (compare this with Latin's development into the Romance languages).
I agree about the intriguing nature of Madagascar. It is a place I would certainly like to visit.
1118. hashke - Jan. 24, 1999 - 1:02 PM PT
check out www.zompist.com/chance.htm for an interesting bit on chance resemblences between languages.
1119. pseudoerasmus - Jan. 24, 1999 - 2:11 PM PT
Hashke (Message #1118)
That's a coincidence, because last weekend when I was leafing through Luigi Cavalli-Sforza's "The Great Human Diasporas", I came upon his discussion of the linguist Joseph Greenberg's alleged "discovery" of a common root word in all languages: "tik". My first reaction was: this Greenberg guy needs some training in probability theory. And I wondered for a little while whether the binomial distribution might be the appropriate model for chance similarities among languages, but abandoned thinking about it because I didn't know the linguistics. But your link confirms my suspicions.
1120. IrvingSnodgrass - Jan. 24, 1999 - 7:22 PM PT
Hashké:
Interesting link, and the author makes some very good points, but he obviously has little knowledge of Greenberg and Ruhlen's methodology. His points and forays into probability are correct and make mincemeat of comparisons such as the Quechua-Semitic one he offers.
However, Greenberg and Ruhlen have gone far beyond the random occurences mentioned, and have not made the phonological and semantic leaps they are accused of, for the most part. I'll get into this more later.
PE:
I too have problems with Greenberg's "tik." One of the categories Greenberg and Ruhlen look for when ruling out chance similarities is "sound symbolism," including onomatopoeia and words like papa and mama. I think a case could be made for including "tik" in this category. In Indonesian, the word "tik" means "to type," from the sound made when typing. It seems to me that "tik" is a sound which is made by the fingers. It has that sense of a small, tight sound, unlike a sound made by another part of the body. I think this could go a long way toward defining why "tik" and "finger" seem to be related in so many disparate languages.
I don't believe in throwing the baby out with the bathwater, and I don't think this one example ruins Greenberg and Ruhlen's work. I'll go into my reasons (and theirs) in a bit.
I do think they are mistaken, however, in looking for "global cognates," of which they have found only a handful. I think most of these can be ruled out by sound symbolism ("akwa" for water, "tik" for finger). I would place "kaka" for older sibling in the category with mama and papa, as well. That leaves "pal" for "two," which may just be a coincidence, like "ni" meaning "you" in Chinese, Tamil, Swedish, Fijian and Amerind languages.
1121. IrvingSnodgrass - Jan. 24, 1999 - 9:41 PM PT
Let's take a look at methodologies for comparing languages. With living languages, the method is simple: one looks for systematic resemblances. For most groups of languages, this method is very simple and clear. No one argues about the inclusion of any of the 140 languages in the Indo-Hittite (Indo-European) family or any of the 959 languages in the Austronesian family. The relationships between these languages are transparent, as their dispersal has taken place in the past 4000 years or so.
But what of languages which have been separated for longer than this? This is the question Greenberg faced when dealing with the languages of Africa in the 1950s, the languages of New Guinea in the 60s and the Amerind languages in the 1980s. He was able to classify these languages in groups which subsequent investigation has confirmed (with a few minor adjustments).
His methodology is to look at a large group of languages, and to classify them on the basis of systematic relationships into sub-groups by comparing individual languages one-by-one until groupings are established. Only *after* this initial work is done is it possible to make generalizations about a family (or to reconstruct a proto-language).
When languages diverge, sounds and meanings change. Sound shift and semantic shift is well documented. Sound shift is usually systematic, and regular correspondences can be found. For example, from my own personal experience, I've been able to define specific sound changes between Indonesian and Cebuano (Indonesian d = Cebuano l, Indonesian r = Cebuano g, etc.).
[continued]
1122. IrvingSnodgrass - Jan. 24, 1999 - 9:51 PM PT
With languages which have been divergent for any length of time, it is usually impossible to trace vowel shifts. Vowels in any language are much more unstable. The article Hashké linked to makes a big deal about the fact that linguists such as Greenberg rarely take vowels into account, which is unrealistic. In fact, vowels *are* taken into account, but usually with the symbol V (for any vowel). Consonants are ore stable.
For example, if language A has the word "gada" meaning "black," language B has the word "kata" with the same meaning, and language C has the word "agald," it would be logical to postulate that A and B are cognate (devoicing of consonants) while the status of language C is questionable. With a large body of data, such correspondences can be confirmed.
To return to my Indonesian Cebuano example above, I found these words were cognate (among many, many others):
tidur --> tolog “sleep”
beras --> bugas “rice”
If vowels were a consideration, I wouldn't be able to do this. The vowel relationships between Indonesian and Cebuano are not clearly systematic, largely because Indonesian has 6 vowels (i, e, a, o, u, and one I can't recreate on my keyboard, which is a high central vowel) and Cebuano only has three (i, a, o, with o/u alternating allomorphically).
By using the vowel symbol (V), these relationships are clearer:
tVdVr --> tVlVg
bVrVs --> bVgVs
Let's look at two words which are *not* cognate between these two languages, to see how this looks when there is no relationship:
malam -> gabi'i “night
mVlVm -> gVbV'V
There are no points of similarity between these two words.
To find correspondences, we look at all words which appear to be cognate, and then establish the specific relationships (d --> l, etc.), and confirm the relationship between two languages.
[continued]
1123. IrvingSnodgrass - Jan. 24, 1999 - 9:57 PM PT
When finding cognates, these are three kinds of seeming relationships (other than true genetic relationships) which we have to look out for. As Ruhlen says in “A Guide to the World's languages,” p. 11): “Three other possible causes must be entertained, and eliminated, before one may assert that the similarities in question derive from a common origin: chance, sound symbolism, and borrowing.”
Examples of the first category, chance, are German “nass” and Zuni “nas”, both meaning “wet,” or Rumanian “dori” and Lau (an Austronesian language from the Solomon Islands) “dori,” both meaning “to wish for, desire.” These languages are unrelated genetically, and borrowing is extremely unlikely.
Sound symbolism, as mentioned earlier, is onomatopoeia or words like mama and papa.
The most troublesome area is borrowing, and linguists have mistakenly classified languages as related based on large numbers of borrowed words. I'm sure one could make a list of 1000 related words in Spanish and Basque, but the languages would still be unrelated, although their has been a great deal of borrowing. Borrowing usually affects certain lexical domains (such as previously unknown cultural items) and certain grammatical categories (nouns, rather than verbs, rarely inflectional or derivational affixes). Basic vocabulary is fairly resistant to borrowing (eye, nose, ear, head, leg, colors, numbers, pronouns, etc.) -- though not always.
[continued]
1124. IrvingSnodgrass - Jan. 24, 1999 - 9:57 PM PT
All of this is by way of explaining why the criticisms leveled against Greenberg and Ruhlen in the linked article are invalid. Greenberg and Ruhlen are careful about omitting change cognates as well as borrowed words and sound symbolism. When they have postulated that two words are cognate, they are fairly certain that these variables have been controlled for. The article is concerned with the chance occurrences of cognate words, which happens significantly often. Greenberg and Ruhlen are concerned with words which do not occur by chance.
Here is an example. These are reconstructed or attributed words for “hand” from various widely dispersed Amerind language families, as presented by Greenberg:
Penutian “makan”
Central-Amerind “maka”
Chibchan-Paezan “maki”
Macro-Tucanoan “muka”
Equatorial “me'eng”
Macro-Carib “emekun”
Macro-Panoan “moken”
Macro-Ge “mako”
The form mVkV(C) seems fairly widespread. Correspondences of this nature have been found by Greenberg for a large number of words. Sure, coincidence between two languages (or families) may be due to chance, but 8 families?
Another complaint in the linked article about Greenberg and Ruhlen's work is that they take liberties with semantic meanings. This is a common objection I've come across, but every time it seems to come from linguists who aren't familiar with the work itself, and have heard of one example or another from someone else. Unfortunately, the author of the linked article doesn't give any examples. However, Ruhlen gives an excellent example in his book “A Guide to the World's languages,” (pp. 386-387) which I will share with you:
[continued]
1125. IrvingSnodgrass - Jan. 24, 1999 - 10:02 PM PT
“Goddard (1987: 657)criticizes Greenberg for loose semantics, meaning that the semantic connections are implausible. As an example he cites Greenberg's grouping of words meaning “excrement,” “night,” and “grass” in a single etymology. Most linguists would agree that these three meanings have little in common. However, an examination of the etymology itself reveals Goddard's trick, for what he has done is to cite the furthest semantic connections in the etymology while completely excluding the basic core meanings. What we find in the etymology in question are words denoting “dark in color.” In Almosan the meaning is “black” and the overall distribution of meanings in North and South America suggests that this was the original meaning. In Keresiouan we find “dark in color” in Iroquoian and “green” in Keresan. In Penutian the meaning has completely shifted to “green” and its close semantic connections “grass” and “blue.” In South America Macro-Ge preserves the original meaning of “black” in Proto-Ge, while other languages in this group have shifted the meaning to “dirty.” Finally, in the Equatorial group the meaning is “excrement.” As in the previous example, a criticism of Greenberg's book with the allegations of his critics reveals what can only be called a willful distortion of the evidence. It should be noted that Campbell (1988: 600) also criticizes Greenberg's lax semantic standards: “G[reenberg]'s forms are quite permissive in semantic latitude. Semantic equations such as the following are not convincing: ‘excrement/night/grass.'””
[continued]
1126. IrvingSnodgrass - Jan. 24, 1999 - 10:03 PM PT
Two more points about criticisms leveled against Greenberg and Ruhlen:
• Critics such as the one in the linked article often state that Greenberg and Ruhlen pick obscure words from various languages to represent a group. The evidence I've seen shows otherwise, but there is an even more convincing objection to this argument, and that is the work of the Russian Nostraticists (see my recent comments in the International thread). The Nostraticists work exclusively with reconstructed proto-languages in making their comparisons. And they have found even more language families related to Indo-Hittite in their Nostratic group than Greenberg postulates for his Eurasiatic grouping. Eventually, all language families will have reconstructed proto-languages, and I would be very surprised if Greenberg's work isn't confirmed at that time.
• Greenberg and Ruhlen's work in the field of linguistics does not stand on its own. Their work on groups of humans has been confirmed by geneticists such as Cavalli-Sforza and archeologists. There is little doubt any longer that there were three migrations into the Americas, for example: the Amerinds, the Na-Dene, and the Eskimo-Aleut. Is it really surprising that the remnants of the first migration (the Amerinds) share related languages?
It should be noted that any relationships about the phylum level (the 17 phyla, plus 5 language isolates) are at present uncertain. There is little doubt that such relationships exist, but more extensive work is needed to define the various relationships.
1127. IrvingSnodgrass - Jan. 24, 1999 - 10:07 PM PT
I just realized I could have avoided a lot of that blather by making one simple observation:
The article How likely are chance resemblances between languages? is all about chance occurrences of cognates in languages, and Greenberg and Ruhlen control for chance.
1128. DanDillon - Jan. 25, 1999 - 5:31 AM PT
Irv,
Your comments, criticisms and questions strike me as being curiously conservative. In any case, I'd like to know if, in your research into the work of Greenberg and Ruhlen, you've come across a discussion of pidginization. Specifically, when, if ever, does their discussion of sound correspondences evolve into a series of hypotheses on the theory of language change? It seems to me that such a comparative phonological examination would inevitably lead to a diachronic analysis of substratum and superstartum influences in language change. Have you come across such a discussion? Are the two processes even related at all?
1129. IrvingSnodgrass - Jan. 25, 1999 - 5:45 AM PT
Dan:
I've tried to keep my comments conservative, and based on easily-provable facts, to give greater weight to what I'm saying. Perhaps later we can investigate higher-level relationships between languages, as I have done several times previously in the Fray. But before we do that, I'd like to see a sort of consensus on the 17 phyla and the legitimacy of the methodology used. So, the conservative tone has been intentional.
I'm not sure exactly what you refer to with pidginization (or creolization, if you're talking about a stage beyond mere language contact). Ruhlen does comment rather extensively on sound changes, but it all relates to the regular patterns we have come to recognize in linguistics.
I would think that creolization has been a minor factor in the spread of the world's languages, as the genetic data appears to confirm. The number of existing creole languages in the world today is very limited, certainly less than a dozen. Pidgin languages are more noticeable as a force for borrowing, though borrowed words are not considered when evaluating the relationship between languages.
1130. DanDillon - Jan. 25, 1999 - 6:46 AM PT
Irv,
It seems interesting to me that a close examination of pidginization (and its higher-level counterpart, creolization) wouldn't somehow play a role in the study of the spread of the world's languages. Indeed, I am not proposing that pidgins account for the proliferation of the global polyglot, but I am suggesting that a close look at pidginization as a *step in the process*, a stop along the line of the evolutionary trail of language, might be useful in understanding various sound correspondences. As far as I know, this is a rather unorthodox approach, and I hardly have the means to undertake such a lengthy and intricate analysis. However, if we view pidginization, creolzation and decreolization as a process that occurs within the larger context of the evolution and diffusion of languages, we might more fully comprehend certain aspects of the entire sheme of sound correspondences. I do not meant to show disregard for the existing genetic data available to us, but I am suggesting that we could further corroborate the evidence that's out there by a close examination of this "other" evolutionary process. In short, pidgins, creoles and everything in between are not an end, a result of some substratum or superstratum influence, but rather a means to the ultimate establishment of many standardized languages around the globe.
1131. IrvingSnodgrass - Jan. 25, 1999 - 6:59 AM PT
Dan:
Do you have any examples to offer? I've never come across pidginization as a step in language change, and I'd like to know more. Obviously, pidginization leads to a simplification of a language's sound system, and the most elaborate sound systems are found in the most isolated languages (the most elaborate phonemic systems of them all are found in the Caucasian languages), which *may* be evidence of your theory. Tell me more.
1132. DanDillon - Jan. 25, 1999 - 7:19 AM PT
Very well. A basic theory of language change, as I see it, looks like this:
Contact = speaker [+ power] and speaker [- power] >> pidginization
Pidgin = lexicon of speaker [+ power] and grammar of speaker [- power] >> creolization
Creole = lexicon of speaker [+ power] and grammar of speaker [+/- power] >> decreolization
Standard = lexicon of speaker [+ power] and grammar of speaker [+ power]
As I said, this simplified process is merely a step along the way of a much larger and more complicated process. And there are many additional factors that come into play that influence the prospects of a non-standard language's survival (or demise): stability; territoriality; contiguity; and mother tongue nativity. Regardless, it seems to me that the process described above--especially its earlier stages--might likely be part of the explanation for the transmission and the ultimate similarities of two or more langauges.
1133. IrvingSnodgrass - Jan. 25, 1999 - 7:32 AM PT
Dan:
Can you provide historical examples of this process, which can be seen in the existing linguistic data? The process makes sense, but I don't see lasting influence of, say, Dravidian on Indo-Aryan languages, Amerind on Latin American Spanish and Portuguese, or Celtic on English (to give a tiny number of historical examples of language contact), apart from a limited number of lexical items. I'd like to see some evidence of where this process has left some evidence in terms of a simplified or altered grammar or sound system before I buy it as a model for language change.
1134. DanDillon - Jan. 25, 1999 - 7:37 AM PT
As for examples, we may look no further afield than medieval England. I have wondered before (but never in a formal study) if the theory of langage change outlined above might help explain the transition from OE to ME. The Middle English period was the most turbulent period of the linguistic evolution of English; OE was giving way to a much more analytic language. (Open-class morphemes literally exploded onto the scene, to use a colloquialism.) When the Normans first came to England and won the battle at Hastings, Latin, not French, became the language of the law and state. But French was used widely by the court and by the educated population [+ power] across Britain. Generally, French was used to relate urban concepts and modes of living (i.e. its lexicon was appropriated for this purpose), while a transforming grammar (early ME) was used by the rural or country folk [- power]. And since the language of the populus [- power] seems to predict fairly soundly the next language of the state [+ power], a "public language" (literature, media, books), ostensibly an advanced creole at this stage, decided the langauge of the state. Hence, the standard Middle English of Chaucer.
1135. IrvingSnodgrass - Jan. 25, 1999 - 7:56 AM PT
Dan:
I don't dispute the vast influence of Norman French on Middle English, and I think your assumption of this as the defining line between OE and ME is pretty well accepted.
I do, however, take issue with your calling the French-influenced English a "creole." A pidgin (the first step) develops when two cultures are in contact. In this case, the language which would be pidginized would be the +power language, or French. If we were all speaking a form of French today, your theory would work. But we don't speak French.
I can't see where a creole, or even a pidgin French ever developed. And I don't see a pidgin English having developed either, since the French were the +power group. All I see is a large influx of French words and significant influence on English grammar. Rather the opposite of creolization, in fact.
1136. DanDillon - Jan. 25, 1999 - 9:16 AM PT
What is, then, the polar opposite of a creole? An anti-creole? Perhaps I am too fast and loose with my terms (a huge problem in this case, admittedly), but the process is identical to a pidginization, or an anti-pidginization, no? There must be something there. Please pardon my lexical indiscretion.
1137. pseudoerasmus - Jan. 25, 1999 - 9:27 AM PT
I think I missed something. What is the significant influence of French grammar on English grammar?
1138. Jenerator - Jan. 25, 1999 - 9:50 AM PT
Last night, I went out to dinner with a friend, to a tiny Italian restaurant. When we came in, there was a guy slinging pizza dough who greeted us. He said "Hey grab a seat...Malakas". My friend just sat down, not knowing what "Malakas" meant. I was angry, but thought for a moment that I had misinterpreted what he said. So, I walked back over to him and said "Excuse me, did you just say Malakas?" He grinned, and tucked his chin under "Heya, ya know, itsa just for funna". I glared at him, and said "It'sa not so funna whena the people-a you insulta understand whata you a saya, NO?"
I was angry, so we left and went elsewhere.
Why do people do this? It reminds me of all the times my friends have been to Paris only to hear waitors tell them something dirty while they order. Is it a European thing?
Afterall, I've never heard an Hispanic say "Grab a seat.... bandejo" at a Mexican restaurant.
1139. JaDeGoLd - Jan. 25, 1999 - 9:54 AM PT
I believe you mean "pendejo"
1140. Jenerator - Jan. 25, 1999 - 9:56 AM PT
I never did learn how to spell foreign curse words.
1141. DanDillon - Jan. 25, 1999 - 10:06 AM PT
OE was, all things being equal, SVO unless something other than S or V began a sentence. If an adverb began a sentence, then it would be VSO. In subordinate clauses, V came at the end. ME was also SVO. However, the syntax of subordinate clauses also remained SVO, as opposed to the arrangement in OE, which was V-final. This was a major change that came about because subordinate clauses were far less frequently used in ME and thus more vulnerable to change. Basic simple sentences were used most frequently, and so they were least resistant to change. (Syntactic structures that are used most frequently are always less likely to undergo change.) Pardon my prose, but this is cumbersome material.
Basically, ME saw a drastic reduction in word order variations in its evolution from OE. As I mentioned earlier, ME was a far more analytical language than OE. Word order determined the functions of words more so than did any inflectional morphemes as in OE. Phonologically, consonants were simplified in ME, French consonants (/v/ and /z/) were added, and diphthongs were smoothed. Morphologically, there was a reduction of noun cases and ME also lost all vestiges of grammatical gender. There was the loss of dual number in pronouns. Many strong verbs were weakened, and the singular and plural forms of the past tense in strong verbs merged. Needless to say, there were numerous additional changes that took place in the evolution from OE to ME. I've hit the highlights here, and I'm afraid that if I were include any more, dizziness would surely set in.
1142. pseudoerasmus - Jan. 25, 1999 - 10:19 AM PT
But I thought the evolution of English from a synthetic language to an analytic language had begun before the Norman conquest.
By the way, I have never received a satisfactory reply to: why did the Romance languages lose almost all signs of inflection?
1143. pseudoerasmus - Jan. 25, 1999 - 10:23 AM PT
The ninth edition of Sweet agrees with Dillon's description of word order in Message #1141 (which is more or less what modern German has), but also says:
"These three varieties of word order are the basic types; but none of them is invariably observed, and much depends on the weight and rhythm of particular phrases and on the individual style of authors. By Aelfric's time a completely modern order (except for the negative) occasionally appears..."
1144. IrvingSnodgrass - Jan. 25, 1999 - 5:25 PM PT
PE Message #1137:
"I think I missed something. What is the significant influence of French grammar on English grammar?"
It wasn't a case of French grammar having an influence on English grammar. It was more a case of the social upheaval and extensive language contact of the era (especially in light of French being the +power language) being a trigger to changes in English. I think this is the point Dan was describing above. Middle English is much simpler grammatically than Old English. Indeed, English was already on a path to an increasing analytical nature, but the Norman conquest stimulated this process.
Message #1142:
"By the way, I have never received a satisfactory reply to: why did the Romance languages lose almost all signs of inflection?"
Let me start off by saying I am not an expert in any sense of the word on Romance linguistics. The ideas in this post are mere conjecture. Your much deeper knowledge of these languages will be useful in confirming or rejecting my theory.
It's funny, but the situation of the Romance languages may just be confirmation of Dan's Message #1132. When Latin spread out from Italy as the Romans conquered Western Europe, it displaced the local (chiefly Celtic) languages. It is quite likely that this was done through a period of pidginization, followed by creolization, in which the grammar of Latin was simplified. This would explain the loss of inflection. Does this sound plausible?
1145. pseudoerasmus - Jan. 25, 1999 - 5:43 PM PT
"Middle English is much simpler grammatically than Old English. Indeed, English was already on a path to an increasing analytical nature, but the Norman conquest stimulated this process."
Yes, I realise that ME is much simpler than OE, but what I want to know is HOW the Picardo-French of the Normans "stimulated" the process of simplying OE. I don't see it. My understanding is that OE had already greatly "simplified" (began losing inflection and became more modern in word order) long before 1066.
In fact, Dillon's theory exhibits an undue bias toward seeing the Wessex dialect of OE as the base from which ME sprang. (The Wessex dialect was of course the literary dialect and thus the source for most of the historical linguistics of English in that period is based.) However, _The Book of Middle English_, by J.A. Burrow and Thorlac Turville-Petre, has this to say:
"But after 1066, Wessex became no more than one, rather remote, part of a French-speaking king's realm; and the language of Wessex accordingly lots its special status too, ending up eventually as just another form of written Middle Engtish: 'Southwestern'. This development goes a long way to explain why most writings of the 12th century present such a different appearance from those of the 10th or 11th. The language as spoken had, of course, changed in the interval, but the nature of our written evidence for it changed more drastically. 12th century scribes, unlike their Anglo-Saxon predecessors, customarily employed whichever form of English they or their authors happened to use. Hence they represented in their writings changes which had already occured in the spoken language of late Old English, but which had left no more than occasional traces in the writings of that period."
1146. pseudoerasmus - Jan. 25, 1999 - 5:44 PM PT
And we all know that Chaucer's Middle English was East Midland / London dialect, not Southwestern.
1147. IrvingSnodgrass - Jan. 25, 1999 - 5:50 PM PT
PE:
I see the loss of power which English underwent as a major factor in stimulating the simplification of English grammar. Without a standard form, common forms became more popular. A standard form of a language is one of the major barriers to language change. Your example of the loss of status of the Wessex dialect underlines this neatly. Under the French, there was no standard dialect of English. When a standard form re-emerged, it had been greatly influenced by the leveling effect of the loss of power (not to mention a huge influx of French vocabulary).
1148. pseudoerasmus - Jan. 25, 1999 - 5:55 PM PT
In the old language thread, I posted the following comments of Tom Shippey, Prof. of Old English Language & Literature at University College London. It spells out a theory which, by citing the interactions between Danes and Saxons before the Norman Conquest, purports to explain why English lost its system of inflection. Dillon naturally pooh-poohed the idea.
"After Alfred, the Danes and the Saxons lived alongside each other for generations, more or less at peace. Because both their languages had the same Germanic roots, the language frontier broke down and a kind of natural pidgnisation took place that gradually simplified the structure of Old English...Consider what happens when somebody who speaks, shall we say, good Old English from the south of the country runs into somebody from the northeast from speaks good Old Norse. They can no doubt communicate with each other, but the complications in both languages are going to get lost. So if the Anglo-Saxon from the South wants to say (in good Old English) 'I'll sell you the horse that pulls my cart', he says: 'Ic selle the that hors the that hors the draegeth minne waegn'. Now the old Norseman -- if he had to say this -- would say, 'Ek mun selja there hrossit er dregr vagn mine'. So, roughly speaking, they understand each other...they understand the main words. What they don't understand are the grammatical parts of the sentence. For instance, the man speaking good Old English says for one horse 'that hors' but for two horses says 'tha hors'. Now the Old Norse speaker understands the word 'hors' all right, but he's not sure if it means one or two...So: are you trying to sell me one horse or...two horses? If you get enough situations like that there is a strong drive towards simplifying the language."
1149. pseudoerasmus - Jan. 25, 1999 - 5:59 PM PT
I submit that French had little to do with the evolution of English toward being a more analytical language; that the Middle English of Chaucer was in fact an evolution from dialects of OE which had already mostly lost inflection by the time of the Norman Conquest; and that the drastic change in the use of inflection between OE and ME is an illusion caused by the linguists' need to rely on textual evidence to study linguistic evolution.
1150. pseudoerasmus - Jan. 25, 1999 - 6:02 PM PT
Message #1147
To the extent that French had something to do with "picking" which dialect of English became standard, I think it had to do with location. The French seated their government in the East and this naturally became the commercial hub of English. Perhaps this is consistent with Dillon's theory, but I can't tell because his jargon-filled prose gives me a headache.
1151. IrvingSnodgrass - Jan. 25, 1999 - 6:03 PM PT
PE Message #1148:
The quotation nicely fits with Dan's theory of pidginization, and leads to a nice continuum:
1) English-Dane contact: the beginnings of the simplification of English grammar
2) English-French contact: the acceleration of the simplification of English (English changes from OE to ME)
The process of the simplification of English continues today, and I'm sure will be accelerated as English increasingly becomes a world language. Although today we have the stabilizing force of the standard language (unlike during the Norman era).
1152. pseudoerasmus - Jan. 25, 1999 - 6:05 PM PT
By the way, even in the late 12th century, the language of La(y)amon's Brut -- the dialect of the Southwest, the successor to the Wessex dialect of OE -- still exhibits considerable inflection. It certainly has all the cases OE had.
1153. IrvingSnodgrass - Jan. 25, 1999 - 6:09 PM PT
PE Message #1149:
"I submit that French had little to do with the evolution of English toward being a more analytical language"
Not directly, but the combined effects of social upheaval and the loss of a standard language, direct effects of the French conquest, certainly did.
Message #1150:
"To the extent that French had something to do with "picking" which dialect of English became standard, I think it had to do with location."
Yes, of course. Though I wouldn't say there was any "picking." It occurred naturally. The French didn't care much about English, and to the small extent they used it, it was just a tool for communicating with the masses. To that extent, the "picking" of one dialect was really just a coincidence.
1154. pseudoerasmus - Jan. 25, 1999 - 6:11 PM PT
Snirv: There is a reason I put "picking" in quotes. Don't be so literal-minded.
If Dillon's and Shippey's theories are so compatible (and I can believe they are), I wonder why if he was so haughty about it the last time I posted it.
1155. IrvingSnodgrass - Jan. 25, 1999 - 6:12 PM PT
PE Message #1152:
Interesting point. This was probably due to the limited contact the Southwest had with either the Danes or the Normans, compared to other regions.
1156. Msivorytower - Jan. 25, 1999 - 6:13 PM PT
Irving,
sorry to interrupt, but could you check your fraymail account?
1157. pseudoerasmus - Jan. 25, 1999 - 6:14 PM PT
So what is the evidence that French had anything to do with the simplification of English grammar, other than simple supposition?
1158. IrvingSnodgrass - Jan. 25, 1999 - 6:14 PM PT
PE:
Sorry. I was aware why you put picking in quotes. I was just extending the thought (chiefly for those lurking here).
Perhaps Dan has had a change of mind about the influence of pidginization?
1159. IrvingSnodgrass - Jan. 25, 1999 - 6:19 PM PT
PE Message #1157:
"So what is the evidence that French had anything to do with the simplification of English grammar, other than simple supposition?"
The accelerated simplification of English as it changed from OE to ME at the time of the Norman invasion.
Please note that I haven't said that French had any influence on English; rather that *language contact* accelerated the changes.
1160. DanDillon - Jan. 25, 1999 - 6:47 PM PT
"Perhaps Dan has had a change of mind about the influence of pidginization?"
I conceded earlier that pidginization may not, in fact, have taken place with the introduction of Norman French into English, a concession I arrived at by your having clarified for me the + power group and the - power group. A moment of discombobulation, I suppose. Whatever the case, I would be happy to continue this conversation about English changing over time. It is of great interest to me, and my hope is that those who are merely lurking will feel welcome to join in and get their hands dirty (not with mud, but rather with theory, supposition and weighty claims). I do so enjoy participating here myself.
1161. pseudoerasmus - Jan. 25, 1999 - 7:05 PM PT
Message #1144
Your theory of the loss of inflection in the Romance languages does not sound implausible. However, I don't understand why some languages do lose inflection and why others do _not_. Why, for example, has Russian not lost inflection at all? It mains a declension system every bit as extensive as Latin's.
1162. IrvingSnodgrass - Jan. 25, 1999 - 7:23 PM PT
Dan:
Do you think the quote PE offered about language contact between the Danes and the English fits? And do you think the Romance languages fit your pidginization hypothesis?
PE:
I'm not sure why some languages lose inflections while others don't. I'm sure there are linguists who have come up with theories, and I'll see what I can find.
It's clear that language contact alone is not sufficient to make a language lose inflections. Russian (a +power language, with, as you point out, an elaborate system of inflection) has had a long history of expansion through language displacement. On the other side, the Dravidian languages and the Finno-Ugric tongues (-power languages) have kept their elaborate inflectional systems, even as their linguistic position has been displaced. Interesting topic. I wonder if anyone else has thoughts on this.
1163. pseudoerasmus - Jan. 25, 1999 - 7:28 PM PT
Well, Christ, I found in one of my boxes a treasure trove of frightening Old English and Middle English texts from the two electives I took as an undergraduate. (I am certain I was and continue to be the only econ/math major who took an elective course in OE and another in ME.) If any of these texts shed any insight I will report them here.
1164. IrvingSnodgrass - Jan. 25, 1999 - 7:38 PM PT
PE:
Report them anyway! I love that stuff.
1165. IrvingSnodgrass - Jan. 25, 1999 - 7:40 PM PT
Btw, to anyone interested in getting a handle of the history of English, without digging too deep (and without requiring any background in linguistics), I recommend Joseph M. Williams' "Origins of the English language." Clear, accurate and readable.
1166. PincherMartin - Jan. 25, 1999 - 7:53 PM PT
To PE, Snirv, and DanDillon --
This is a fascinating conversation. Keep it up.
1167. pseudoerasmus - Jan. 25, 1999 - 8:16 PM PT
Well, the brief perusal of these OE & ME texts was pretty useless. I had in mind to compare OE texts written in non-Wessex dialects with ME texts written in the East Midland dialect, to preliminarily corroborate my thesis that the OE ancestor of Chaucer's ME had already lost most of its inflection before the Norman Conquest. This was meant to counter Snirv's contention in Message #1159 that "the accelerated simplification of English as it changed from OE to ME at the time of the Norman invasion" was evidence that French had anything to do with the simplification of English grammar. But since the OE texts I have are for literature courses, not linguistics courses, they pay little attention to linguistic variety. Only Sweet offers some fragmentary texts in a non-Wessex dialect of OE -- "medicinal recipes" -- but these are woefully useless for my purposes. All the same, I still don't see any evidence that French accelerated the simplification of OE into ME.
1168. IrvingSnodgrass - Jan. 25, 1999 - 8:21 PM PT
PE:
Would you agree that the lack of a standard form of a language will accelerate language change?
1169. pseudoerasmus - Jan. 25, 1999 - 9:34 PM PT
Message #1168: Yes, but Old English lacked a standard form, in the sense that there were many competing dialects. It's just that unlike the dialects of Middle English, not all OE dialects produced a major literature. Only one did.
1170. pseudoerasmus - Jan. 25, 1999 - 9:37 PM PT
In Message #1162 Snirv mentioned that Dravidian languages are still inflected. Well, surely, the Indo-Iranian languages are a vast field of data from which evidence on how languages maintain or lose inflection can be gathered!
My limited input:
Urdu (and presumably Hindi) maintains a limited system of declesion. The cases include the direct or nominative; the vocative; and the "oblique", which is used when the noun is part of a prepositional phrase, except that in Urdu postpositions are used. Thus, "kamrá" [room] is in the nominative case, but "kamre mé" [in the room] is in the oblique. (I use the accent to indicate a long vowel.) In some circumstances the direct object of a verb takes the nominative but in others the oblique case is used. Adjectives are similarly declined.
Urdu has no genitive case per se, and possession is indicated by the use of the oblique with the postposition "ka". (Strangely, this postposition agrees in number and gender with the object possessed -- which does not occur in any other language I know of. Elaboration on this issue provided upon request.)
Persian/Pushto is almost uninflected. There are no case declensions; the plurals are formed with no problem; and while Persian has no gender even, Pushto has the trinity of masculine, feminine and neutral, a remnant, I assume, of older practises in the Iranian language family.
Punjabi has the equivalents of nominative, genitive, accusative and dative cases. Nouns, pronouns and adjectives are all declined accordingly.
1171. IrvingSnodgrass - Jan. 25, 1999 - 9:46 PM PT
Inteersting, PE. I have little knowledge of modern Indo-Iranian languages, but I am aware that Sanskrit had a highly developed system of inflection (the linguistic term is case system, btw, but we don't need to use it). So it seems that modern Indo-Aryan languages have simplified the system, although they haven't lost it. This may be due to language contact, although most of the languages Indo-Iranian presumably displaced were probably Dravidian, with their own highly developed systems of inflection. It's a very good example to add to the pattern, though.
1172. IrvingSnodgrass - Jan. 25, 1999 - 9:52 PM PT
I meant to add that Sanskrit had a case system similar to Latin, so comparing the situation of the Romance languages with the modern Indo-Aryan languages is a good comparison. Perhaps the difference lies in the existence of complex case systems in the displaced languages of India (Tamil's is, if anything, *more* complex than Sanskrit or Latin), as opposed to a lack of similar systems in the displaced languages of Europe (Celtic was rather unique among the ancient Indo-European branches in that it lost its inflections early on).
1173. pseudoerasmus - Jan. 25, 1999 - 10:01 PM PT
What I don't understand is why the modern German declension system has changed so little from Middle High German.
I have been told that Old German was a fully inflected language, and since Middle High German is only partially inflected, you would think the trend toward the loss of inflection would continue in modern German. But no.
In modern German as in Middle High German, nouns are minimally declined, but adjectives and pronouns are fully declined. There aren't even that many differences in the word endings between the modern and the medieval forms.
In modG, for the adjective "gut", the (strong) endings for masculine, feminine and neuter are:
NOM er, e, es
GEN en, er, en
DAT em, er, em
ACC en, e, es
In MHG,
NOM er, iu, ez
GEN es, er, es
DAT em, er, em
ACC en, e, ez
In some ways, ModG inflection has become _more_ complex than MHG inflection. Both languages have "strong" and "weak" endings for declining adjectives. But modern German adds "mixed" endings.
1174. IrvingSnodgrass - Jan. 25, 1999 - 10:12 PM PT
PE:
My guess is that modern German has become "frozen" in its standard form, much like English. The pace of language change has slowed down considerably in European languages since widespread literacy has become common. A standardized language with widespread literacy makes it that much harder for language change to occur (though it still occurs, obviously).
1175. pseudoerasmus - Jan. 25, 1999 - 10:32 PM PT
Message #1171 & #1172
• Yes, whenever I use the word "inflection", I invariably mean a case system or declension system by which nouns, pronouns and adjectives are modified by word endings. I don't mean verbal inflections or conjugations.
• tell us something about the case system of Tamil or any other Dravidian language you know.
• why would Iranian languages (at least Farsi and Pushto) almost completely lose their case systems -- assuming that their common ancestor had a complex case system, which I assume though I don't really know.
1176. pseudoerasmus - Jan. 25, 1999 - 10:34 PM PT
A few other notes:
(1) Old French (contemporaneous with Middle Engtlish) and medieval Provençal had a two-case system of nominative and "oblique", but only the definite article was declined.
(3) Modern Russian has six cases (with three genders), while its ancestor Old Church Slavonic had seven.
(4) Does anyone know whether modern Greek has a case system? Ancient Greek had a case system of nominative, genitive, dative, accusative and vocative -- a smaller numer than Latin or modern Russian.
(5) Dillon can tell us whether modern Arabic has a case system, for I have no idea, but Classical Arabic certainly had it: nominative, genitive, accusative, vocative, and oblique (prepositional).
1177. pseudoerasmus - Jan. 25, 1999 - 10:40 PM PT
Oh, the modern Russian cases are nominative, genitive, accusative, dative, instrumental (the equivalent of the Latin ablative), and locative (to indicate place). Old Church Slavonic had all these plus a vocative.
Latin of course has nom, gen, acc, dat, ablative and voc.
1178. IrvingSnodgrass - Jan. 25, 1999 - 10:51 PM PT
PE:
"• Yes, whenever I use the word "inflection", I invariably mean a case system or declension system by which nouns, pronouns and adjectives are modified by word endings. I don't mean verbal inflections or conjugations."
Yes, I had assumed that we were speaking of the same thing.
"• tell us something about the case system of Tamil or any other Dravidian language you know."
Sure. I have to do some real world work right now, but I'll post on this later today (well, my day, anyway).
"• why would Iranian languages (at least Farsi and Pushto) almost completely lose their case systems -- assuming that their common ancestor had a complex case system, which I assume though I don't really know."
I would guess (again, this is a guess) that it may be because they didn't have a language like Sanskrit keeping them in line over the years (though that didn't help with Latin and the Romance languages). Avestan probably had a case system, though I don't really know. I'm quite certain proto-IE had a well-developed case system, since it shows up in such similar form in so many branches. I've never really thought about why languages lose case systems (nobody ever seems to *develop* them at a later date), and this is a very intriguing topic.
I'm not sure about case in Modern Greek, but I know where to look it up, and I will.
Back in my linguistic study days, I studied the case systems of Tamil, Latin, Sanskrit and Finnish in some depth. The most complex of these were Tamil and Finnish (12 and 10 cases, respectively, if memory serves me), though Latin and Sanskrit had very similar and well-developed systems. I'd have to look at my sources to refresh my memory, though.
1179. IrvingSnodgrass - Jan. 25, 1999 - 11:00 PM PT
PE:
Modern Greek has four cases (Nominative, Accusitive, Genitive, Vocative). The only additional case which Ancient Greek had was dative.
I've found some great stuff on Tamil, which has 8 cases which differ in their singular and plural forms, making 16 cases in actual practice. Lots of other interesting things, too. I'll post it later.
1180. pseudoerasmus - Jan. 25, 1999 - 11:26 PM PT
Message #1179
"Tamil...has 8 cases which differ in their singular and plural forms, making 16 cases in actual practice."
You mean you're going to count plural forms of the word endings as separate cases? Well, then, for adjectives, you should also count gender separately too. Ancient Greek has five declensions, three genders and three numbers (sing., plural and dual), although many permutations for the dual as well as the vocative are missing. So the declension paradigm for the typical adjective in Ancient Greek contains 30 different forms.
1181. alistairconnor - Jan. 26, 1999 - 3:02 AM PT
Are there any other Germanic languages, apart from English, which are not inflected?
Pseud, Message #1116: Malgache women are the most beautiful in the world. They also have a most attractive manner of speaking French. More excellent reasons for a visit.
1182. stostosto - Jan. 26, 1999 - 3:32 AM PT
alistairconnor
Pseuder asked me once if the Scandinavian languages have lost their inflection. I had to have explained what that meant, which he did by giving an example from German - it has to do with grammatical cases, ain't that right? On that basis I concluded that the Skandi lingui are non-inflecti. (I am a Dane, thus my angle).
I also stated that the horrors of German grammar have contributed non-negligeably to its decline as a first foreign language in Scandinavia. To which he said something like "hahaha - German grammar is child's play compared to really seriously inflected languages for grown ups - like Russian".
He also puzzled me by stating - on the basis of a tiny example of Skandi lingo I offered - that we have "rudimentary gender". It sounds like an insult, which I don't think it is, but I have been wondering ever since just the same...
1183. jkuzmak - Jan. 26, 1999 - 3:32 AM PT
Simply want to say that I really appreciate this thread and intend to read the whole thing sometime.
1184. pseudoerasmus - Jan. 26, 1999 - 3:43 AM PT
Message #1182
"He also puzzled me by stating - on the basis of a tiny example of Skandi lingo I offered - that we have 'rudimentary gender'. It sounds like an insult, which I don't think it is, but I have been wondering ever since just the same..."
Well, the Economist knew what I was talking about. In the 23-29 January issue, which contains a "Survey of the Nordic countries", there is a photo of one whom I take to be Stostostosto, lying totally nude in snow...with a rather rudimentary gender.
[Sorry, I've had only 3 hours of sleep.]
1185. DanDillon - Jan. 26, 1999 - 5:42 AM PT
The scenario outlined in the quote in Message #1148 does indeed seem highly plausible. In fact, I consider it a very clear (if not over-simplified) explanation of how language contact accelerates language change. Also, I have no memory of ever discounting such a quote; it was a cockeyed definition of the Danelaw to which I objected.
1186. DanDillon - Jan. 26, 1999 - 6:05 AM PT
The Arabic that I speak is not exactly what you'd call standard, by any stretch, but Derija, the colloquial Arabic of N. Africa, does have a case system. Just running it through my head--I've had very little formal training in the language--I believe there are four cases: nominative, accusative, genetive, and ablative. Perhaps there is a vocative case as well(?). Somebody please correct me here if I am wrong.
On an etirely different topic, there were some very interesting consonantal changes that occurred during the transition from OE to ME. Aside from the voiced fricatives /v/ and /z/ entering the language from French, word initial sequences hl-, hn-, and hr- (OE hleapan "leap," hnutu "nut," hraðor "rather,") were simplified to l-, n-, and r- (ME lepen, nute, rather). Also, initial hw- became wh- (OE hwæt > ME what) with the moderately aspirated /h/ dropping out in many dialects. Similarly, the word-initial wr- cluster simplified in its pronunciation to /r/ (write, wrought). These changes took place, of course, over many years and didn't catch on at all in certain regions until the late ME period.
One last fun fact: Present participial forms, largely missing in OE, gained currency in ME. The -ing siffix derived from the OE noun ending -ung (OE lëornung > ME learning)--hence, the ModE gerund.
1187. IrvingSnodgrass - Jan. 26, 1999 - 7:08 AM PT
PE Message #1180:
No, I wouldn't count plural forms as separate cases, but they do add more case forms to a language. My study of case today has shown me that plural cases are not uncommon, making a real headache for learners of a language. In the examples I will give, I will include plural forms where available, but they are not separate cases. This has been a most interesting refresher on case, btw. Thank you for directing me to it.
Dan:
Standard Arabic has three cases, Nominative, Gentive and Accusative, indicated by short vowel suffixes:
kitabun "book" (Nom.)
kitabin "book" (Gen.)
kitaban "book" (Acc.)
Next: an overview of case in a few languages.
1188. IrvingSnodgrass - Jan. 26, 1999 - 7:14 AM PT
Here then is a look at the case systems of a few languages. I'm going to keep things simple, to save time and effort. I will, however, be available to answer specific questions on any of these languages or their case systems.
All the information below (as well as the info on Greek and Arabic above) comes from the excellent book "The World's Major languages," edited by Bernard Comrie (Oxford University Press), which provides a comparative grammar and phonology of the world's major languages and language groups, including detailed studies of the world's 37 most important languages (it was a fascinating, and (to my mind) successful effort to determine just what the world's most important languages are).
[continued]
1189. IrvingSnodgrass - Jan. 26, 1999 - 7:24 AM PT
oops... before I get atarted, one more answer.
Alistair:
Dutch nouns are not inflected any more than English is (i.e, only for Genitive), nor are adjectives inflected. Given Stostosto's comments on Scandinavian, I conclude that only German retains a case system (basically only observable in adjective forms) among Germanic languages.
Old Germanic languages all had 5 cases , though these differed by language. Gothic and Old Norse had Nom., Gen., Dat., Acc., and Voc. Old English, Old Saxon, and Old High German had Nom., Gen., Dat., Acc. and Inst.
1190. IrvingSnodgrass - Jan. 26, 1999 - 7:46 AM PT
We'll start our survey of case with Indo-European languages. I'll be using the following abbreviations for cases (with simple (and not entirely accurate) explanations for the linguistically untrained):
N = Nominative (subject form)
A = Accusative (object form)
G = Genitive (possessive form)
D = Dative (the same meaning as the preposition "to")
V = Vocative (addressing someone/thing)
I = Instrumental (instrument)
L = Locative (location)
Ab = Ablative (the same meaning as the preposition "from")
Other cases I'll present in full form. For some languages, I have forgotten what the cases mean(!)).
Ancient Greek had 5 cases (N, A, G, D, and V). Modern Greek has lost the D form, but retains the other three. I would provide examples, but Slate won't accept Greek letters, and I don't know enough Greek to romanize the letters. In both Ancient and Modern Greek, the forms differ for Gender and stem (a-, o-, or consonant).
Latin had 7 cases, and is described as being typologically close to proto-IE. Latin cases were organized in six paradigms. These paradigms tended to be the same for adjectives and nouns (with two exceptions), and reflected proto-IE paradigms with one exception. For more detail, see Comrie's book.
Because of format limitations here, I can't present all the paradigms, so I will present one: the paradigm for popul-us 'people':
N popul-us
V popul-e
A popul-um
G popul-i
D popul-o
Ab popul-o
L no form for 'popul-," but compare hum-i 'on the ground)
N, V pl. popul-i
A pl. popul-os
G pl. popul-orum
D, Ab, L pl. popul-is
1191. IrvingSnodgrass - Jan. 26, 1999 - 8:00 AM PT
To look at one more classical IE language before I move on to modern languages, here is the breakdown for the eight cases found in Sanskrit for the word deva 'god':
N deva
V devas
A deva
I devam
D devena
Ab devaya
G devat
L devasya
Dual forms:
N/V/A devau
I/D/Ab devabhyam
G/L devayos
Plural forms:
N/V devas
D/Ab devebhyas
A devan
G devanam
All eight forms are distinguished in the singular masculine -a stems (as in the example). Endings differ by gender and stems, as in other IE languages. The paradigms are less complex for other genders and stems (but not much). For more info, see Comrie.
1192. IrvingSnodgrass - Jan. 26, 1999 - 8:20 AM PT
I might as well start modern IE languages with Hindi-Urdu, the direct decendant of Sanskrit. Hindi/Urdu has two numbers, singular and plural (note the loss of dual forms); two genders, masculine and feminine (as in Sanskrit); and three cases, direct, oblique and vocative (a significant change from the Sanskrit system). As far as I can tell, the oblique and vocative forms are always identical, so it's more like two cases. For information on how these cases operate, see PE's informative and accurate Message #1170. Here are a few examples:
Masculine nouns ending in -a:
Dir Sg larka 'boy'
Obl Sg larke
Voc Sg larke
Dir Pl larke
Obl Pl larko
Voc Pl larko
Note: The first vowel in "lark-" should actually be a schwa.
Masculine nouns ending in -i:
All Sg forms mali 'gardener'
Dir Pl mali
Obl/Voc Pl maliyo
Masculine nouns ending in -u:
All Sg forms sarhu 'wife's sister's husband'
Dir Pl sarhu
Obl/Voc Pl sarhuo
Masculine nouns ending in a consonant:
All Sg forms nokar 'servant'
Dir Pl nokar
Obl/Voc Pl nokaro
Note: the "a" in 'nokar-' should be a schwa
Feminine nouns ending in -i:
Dir Sg larki 'girl'
Obl Sg larki
Voc Sg larki
Dir Pl larkiya
Obl Pl larkiyo
Voc Pl larkiyo
Note: The first vowel in "lark-" should actually be a schwa.
Feminine nouns ending in -a:
All Sg forms mata 'mother'
Dir Pl matae
Obl/Voc Pl matao
Feminine nouns ending in -u:
All Sg forms bahu 'daughter-in-law'
Dir Pl bahue
Obl/Voc Pl bahuo
Note: the "a" in 'bahu-' should be a schwa
Feminine nouns ending in a consonant:
All Sg forms bahan 'sister'
Dir Pl bahne
Obl/Voc Pl bahno
Note: all "a's" in 'bahan' and 'bahn-'should be schwas
1193. IrvingSnodgrass - Jan. 26, 1999 - 8:28 AM PT
Before I continue with a survey of case systems, there is another message above I want to respond to:
Dan Message #1185:
"it was a cockeyed definition of the Danelaw to which I objected."
This reminds me of a policy BG and I enacted for our Cyberdeli here in Bali. We discovered that many of our customers were Europeans, and they were unable to understand much of our written instructions or information (including advertising), especially that which we had prepared with a bit of humor in mind. Since most of these Europeans for some odd reason are from Denmark (70% of our customers are Danes, in fact), we decided we would make a rule that we would write nothing which the average Dane couldn't understand... keeping everything simple. What else could we call this rule but the "Dane Law"?
1194. DanDillon - Jan. 26, 1999 - 8:35 AM PT
Message #1193
Nice one, Irv. I could almost hear the rim-shot.
1195. IrvingSnodgrass - Jan. 26, 1999 - 8:42 AM PT
To return to a previous question about inflection in Iranian languages, I found this quote from Gernot L. Windfuhr in Comrie (p. 530):
"Unlike East Iranian Pashto and many smaller dialects, [Farsi] has almost completely lost the inherited synthetic nominal and verbal inflection and their inflectional classes, and thus the inflectional distinction of case, number and gender as well as of tense, mood, aspect and verbal gender."
I guess that settles the question about Farsi. As for Pashto, D.N. MacKenzie atates the following in Comrie (p. 554):
"Although it has departed considerably from the morphological patterns of Old and even Eastern Middle Iranian..., Pashto has nevertheless a remarkably complex nominal and verbal morphology. Two grammatical genders (masculine and feminine) and two numbers (singular and plural) are distinguished in noun and, in part, verb. Although the nominal case system has in part been reduced to a contrast between direct and oblique, there is in the singular also a vocative and a second oblique case used in conjunction with certain prepositions."
1196. pseudoerasmus - Jan. 26, 1999 - 8:49 AM PT
Snirv
(1) A slight correction. In Urdu/Hindi the oblique and vocative cases are not the same in the plural. Whereas in the oblique plural, the case ending vowel O is nasalised, the case ending vowel O in the vocative plural is _not_.
(2) I think your comments on Greek and Latin are incomplete without stating that in both languages, nouns and adjectives are classed according to the system of declensions they belong to. Thus, such things as "first declension nouns" as well as "third declension nouns". Latin has five classes of nouns and Greek, three. And the word endings are completely different the one from the other. Just compare your paradigm for "populus" with the one for "miles" (soldier):
NOM miles
VOC miles
ACC militem
GEN militis
DAT militi
ABL milite
Of course the -is of "militis" is short, whereas as the -is of "populis" is long, so the two endings are not the same.
(3) If I weren't so sleepy from the little sleep I had last night, I might be persuaded to upload to notebook website a complete declension paradigm for a Greek adjective. But I doubt anybody would be interested.
1197. IrvingSnodgrass - Jan. 26, 1999 - 8:56 AM PT
I have one more Indo-European language to present before I proceed to the two languages which I'm most interested in here (yes, all of this is merely a build-up for those).
Russian has preserved much of the case system of classic IE languages. There are 6 cases, and forms differ for number, gender and stem endings, as in the classic languages presented above. In fact, of the major modern IE languages, Russian has preserved more of the proto-IE case system than any other language. I'd be interested in seeing the systems in Baltic languages, however, since I understand they are the most "archaic" of all IE tongues. Unfortunately, I have no data here on Lithuanian or Latvian.
The cases of Russian, presented for a-stem words only (they differ for masculine o-stem, neuter o-stem, and i-stem), is as follows:
Singular:
N straná
A stranú
G strany
D strané
I stranóy
L strané
Plural:
N strány
A strány
G stran
D stránam
I stránami
L stránax
1198. DanDillon - Jan. 26, 1999 - 9:00 AM PT
Where does this information come from? Cite sources, please.
1199. pseudoerasmus - Jan. 26, 1999 - 9:08 AM PT
Message #1195: I think this may be a dialectical difference, but according to Shaista Wahab's _Pashto Grammar and Reader_, there are three genders in Pashto. The grammatical cases are direct/nominative, vocative and just one oblique.
1200. pseudoerasmus - Jan. 26, 1999 - 9:09 AM PT
Message #1198: Snirv has already cited his source.