B. Lukács
President of the Matter
Evolution Subcommittee of the Geonomy Scientific
Committee of HAS
CRIP RMKI, H-1525 Bp. 114,
Pf. 49,
lukacs@rmki.kfki.hu
ABSTRACT
S. M. Stirling is
writing books in Alternate History. Besides the obvious evolutionary
connections, these books suggest lots of interesting scientific or scholarly
problems, some discussed in the books, some not. Here I discuss one such
recurrent in his
0. PROLOGUE
S. M. Stirling is
producing a big variety of Alternative History sci-fi’s. Alternative History is a subclass of sci-fi’s, where the fantasy does not enjoy its unlimited freedom. In an
ideal AH story only one event has other outcome than on our timeline (that is
the PoD, Point of Divergence), especially an event
which "might have happened as well otherwise", by other words, where
even now we do not know if the particular outcome was necessary or not. (Here I
omit philosophical discussions of Free Will, determinism/indeterminism, the
Everett Theory of Quantum Mechanical Measurement and such, for simplicity.) So
at PoD one event has another outcome, and then the
author continues the story, with as
strict "causality" as possible. Of course the PoD
acts then as an initial condition different than in Our TimeLine (OTL). Then
the author elaborates this Alternative History.
Now, we do not yet
know if such concurrent timelines can be realised "synchronously", or
not. (Against the a priori
impossibility of the alternative see
The machinery can be
demonstrated by means of, e.g., Poul Anderson's
works. In Eutopia [2] a modern (c. 1970 AD) Greek
social researcher from a timeline where Alexander III of Macedon did not drink himself exactly into death but
after some time recovered, so Aristotle
did not have to leave the Lyceum in 322 AD, can switch timelines,
and explores one where Alexander did die prematurely, but the Tours-Poitiers battle in 732 AD
had an outcome opposite than in OTL. So the Frankish Kingdom became weaker, the Pippin-Charlemagne line could not
substitute the Merovings, so in the next 200
years Dane Vikings & Magyars destroyed the Western Christian civilisation.
In 1970 AD Danes can be found in Canada & Minnesota (
S. M. Stirling works
in the AH scheme. In one sequence (Lords of Creation [4], [5]) the PoD was 200 My in the past; a
superhuman civilisation terraformed Venus & Mars,
and from time to time they imported new animals & plants to the other two
planets. Terrestrial history was practically not influenced until 1947 AD, when
astronomers detected the habitable planetary surfaces. Therefore Cold War went
to space. The Draka sequence of Stirling puts the PoD to somewhere at the second half of the XVIIIth century, Northern American
loyalists go to South Africa, so the XXth century
will be very, very different. (But this sequence is par excellence Military SF,
so I ignore it completely.)
Now, Stirling &
Flint invented a somewhat different new AH scenario. (I think
The idea is
"simple". A tremendously superhuman ET civilisation sends back a
negligible part of our contemporary Earth into the past, and replaces the
surface with the past one. In
While this opus is not
yet comparable in size and details with that of Honoré
de Balzac, it may be later; and while Balzac had to know only his present society and its immediate past, Stirling
is confronted with scientific problems, and with a number of different
societies. True, the scientific problems are not direct. Earthpeople
do not understand the Lords of Creation; not even the bioscience of Martian
hominids. But
So far he generally
fulfils the expectations. But the opus is a challenge for somebody (I mean myself)
who is the President of a Matter Economy Subcommittee. Evolution of sciences,
evolution of languages, evolution of hominids...
I am going to make a
series of comments of that opus. That
will not be criticism. Nobody is
interested if I like or dislike a book. Rather I am going to discuss different
items in the books from the viewpoint of the actual science (or, sometimes,
scholarship).
Sometimes this is
really a challenge. E.g. Lords of Creation definitely use a
physics unknown for us. And our present physics is not the Physics. For 2000 years it was not only elementary
experience but also Fundamental Theory that there is no Motion (at least,
except for transients,) without Force. People generally believe that Aristotle
was stupid and the later generations did not dare innovate; but Aristotle did
know that his theory was not yet complete. See his own words about the problem
of ballistics (spears and arrows) [16]. In Chap. 32 of Mechanics he writes:
"Why is it that an object which is thrown eventually comes to a
standstill? Does it stop when the force which started it fails, or because the
object is drawn in a contrary direction, or is it due to its downward tendency,
which is stronger than the force which threw it? Or is it absurd to discuss such questions, while the principle escapes
us?" (Italics are mine.) The paragraph is clear enough even it
contains only questions. The Stagirite did know that
ballistics was at best very difficult to describe even in his theory; and he
was not able to do. Still, missiles flew; and later somebody would improve even
his description.
Before Einstein's
success not only everybody in physics was convinced that Flow of Time was
independent of spatial motions; everybody believed that even a doubt would be
meaningless. Special Relativity appeared a mere 103 years. Similarly, until the
formulation of Quantum Mechanics (more or less 1926) physicists were convinced
that any object has its sharp momentum and
position, although maybe we do not know them; and similarly that one object
cannot be simultaneously here and there. It seemed defetism
to question the first statement and unscientific magic the second. That was 82
years ago.
Our present science
will be so primitive and untrue for the XXVth
century as the Aristotelian physics for us. Still, Aristotle was a true
scientist (while Platon, Marcion,
Plotinos and Melanchton
were not), as a true physicist as e.g. Galileo,
In the first study I
reflect to the opinion suggested by [7]-[9] (most explicitly in Chap. 7 of [7]
and Chap. 26 of [8]) that agglutinative languages are
hopeless to learn, at least for Indo-Europeans. I do not believe this; in the
thousand years of the Regnum Hungariae many Germans,
Slovakians, Rusyns, Roumanians
and Serbs learnt agglutinative Magyar, the language of the central parts of the
country. Not everybody spoke it, and some spoke not too stylishly; but
multiplicity of nominal cases, more than one endings simultaneously attached to
a root, Vowel Harmony and such were never insurmountable barriers.
Indo-European languages as Lithuanian or Hittite are indeed somewhat simpler;
but agglutinative languages are not absurdly complicated.
1.
In [7]
Indeed, she is able to
start some language course with a rescued castaway; although I simply cannot
believe that an astronomer would try with Pl2 in "Do you understand?"
as "Ar mane spurantate?" instead of Sg2. But
the Nantucketers give up halfway the effort to work
out understanding; instead they hire a Tartessian
merchant, slaver &c. who speaks some Mycenian,
good proto-Celt and a little of the other language of
The Tartessians are, of course, early ancestors of Basque, so agglutinative. And the aborigines of
Merry Old Britain are also speakers of an agglutinative
language. Agglutinative languages are rather hopeless (at least for
Now, in Chap. 2 I
discuss the meaning of agglutinative,
while in Chap. 3-5 I give a demonstration about noun, adjective, numeral and
pronoun agglutination from the most agglutinative language of the
2.
WHAT IS AN AGGLUTINATIVE LANGUAGE?
Modern linguists do
not like too much the term agglutinative;
the overwhelming majority of them is not the speaker of Magyar. True, the
definition, from a century in the past, was not quite strict. Still, it was
useful; you may, e.g., follow [17]. If I try to discuss the problems about Swindapa Kurlelo's agglutinative
language, then a classic work from the first half of last century is just
adequate.
I am not following exactly [17], but you can find every
necessary definitions there.
There are isolating, agglutinating, flectating and polysynthetic
languages, with all the possible transitions as well.
Isolating languages in the ideal case put together unchanged blocks ("words").
The par excellence isolating language is (Mandarin) Chinese; it can be written
with pure ideograms since the blocks are unchanged. (Japanese uses the same
ideograms, but not being isolating it needs syllabaric
signs as well.) Practically English and Afrikaans are also isolating languages;
they almost totally gave up the original grammatical endings. (And search
engines are much less capable in non-isolating languages.)
Flectating languages can be
demonstrated by the older Indo-European languages. Latin is extinct and Italian
lost the various endings of nouns. (The system kept very much of the verbal flectating
forms, but I
ignore verbs in the whole study,
for simplicity.) Practically the only thing you can do with an Italian noun is
to form its plural: Italiano>Italiani,
but Italiana>Italiane.
However the old system is hardly simplified in the modern Slavic & Baltic
languages. (For this reason is the half-Lithuanian astronomer included in
[6-8].) The Common IE language had 8 (or 9?) different nominal cases both in
Singular and in Plural (the system was rudimentary in Dual), and Lithuanian
still keeps 7 as:
|
Nom |
arklys |
horse |
|
Gen |
arklio |
of horse |
|
Dat |
arkliui |
to horse |
|
Acc |
arkli |
horse1 |
|
Ins |
arkliu |
with horse |
|
Loc |
arklyje |
at horse2 |
|
Voc |
arkly |
horse! |
1 The difference of Nom and Acc is "I" and "me".
2 Or "on", or even "in".
Table 1: One of the 5
Lithuanian nominal declensions
The above Lithuanian
example shows that the last 1-2 syllables of the noun changes according to the
case (here: -ys, -io, -iui, -i, -iu, -yje,
-y). This is declension. Generally
different declensional schemes coexist; in Lithuanian five, with further subclasses.
I must confess that I omitted an "accent
mark" in Acc. Hittite had one more case, the Ablative, sometimes kept in
Latin, sometimes not. In Latin the system had been just degenerating when it has
become frozen. It is in principle a 7-case system, but the Vocative differs
from the Nominative only for 1 declensional class of the 5 (Brutus – Brute) and
Locative is preserved only in topographic words (Carthago
– Carthagini). Anyways, flectating
languages do have good points. Obviously it gives a
flexibility to a language if the different roles in the sentence are
characterised by different forms. E.g. then the word order is more or less
free, so it can be used for such purposes as emphasis, stress, &c. For
English speakers this may be “complicated”, but look, modern Slavic languages
preserve almost a Lithuanian-sized “complication”, even if in them the Vocative
tends to coincide with the Nominative and Accusative sometimes either with Nominative
or with Genitive. Every Slavic child learns a 6-case nominal paradigm in 3
years.
Polysynthetic languages can mark "more than one"
grammatical roles in one word, e.g. that the subject is singular or plural, and the object is singular or plural.
This is "too much", at least to IE speakers.
Agglutinative languages do not decline
the noun according to cases, but put together unchanged roots and unchanged
endings. (So they glue together
blocks.) This is of course only the ideal situation; in living languages one
can expect slight changes for easier pronunciation and such; but the rules are
simple & unequivocal. So if the language borrows a word from somewhere,
everybody can put that word to any grammatical role.
To demonstrate what
words look in an agglutinative language and how the system works, I take a Karaim example Karaim is an
Eastern Kuman or Kipchak
language, spoken in Lithuania & Crimea, and (besides of Hebrew letters) in
Lithuania uses Latin letters (with more or less Lithuanian orthography).
Turkish languages are rather strict in agglutinative rules. Karaim
is near to the language of the
The
expression: “for our doctors” is a single word (of course) in Karaim (as too in Ottoman Turkish & Magyar), but,
contrary to Ottoman Turkish, lacks any international loanwords: savuchturuvčularymyzdan.
It is built from blocks: sav-uch-tur-uv-ču-lar-ym-yz-dan.
Sav- is the
root, roughly “health”. (In Ottoman Turkish “sağ”
= “right”.) The block –uch
forms a verb, “get healthy” or “be healthy”. Then –tur makes the verb causative, (in
Ottoman Turkish also –tur, more precisely –tur/-tür/-tır/-tir, a 4-form suffix, but that is Vowel
Harmony, see Appendix),
so savuchtur
is “to cure”.
Then –uv (in Ottoman Turkish –u)
forms nouns from verbs, and –ču (Ottoman Turkish –cu) is the person of the cure, so savuchturuvču
= “doctor”. (And this is the reason that I started with Karaim,
not the Ankara Turkish: in the latter savuchturuvču
= hakım, from Arabic.
Now, -lar makes Plurals, -ym marks the
possession of First Persons, but the affixed –yz makes the possessor Plural;
and finally –dan
is simply the Ablative ending (the last four endings are exactly the same in
In Magyar it is “orv-os-a-i-nk-tól”,
slightly shorter. A good Magyar example (albeit not so overwhelming as the Karaim one, is a word we tell when drinking (so cca. “Cheers!): Egészségedre! So “To thy health. Here “Egész” is
“whole”. The suffix “-ség” makes an abstract noun, so
“egészség” is “health”. The suffix “-d” marks the Sg2
possessor, and “-re” is “to, onto”.
My guess is that Turkish languages have the strictest rules
of agglutination. But for quantity maybe
Manyshi has the greatest number of agglutinative
verbal endings in the
People
generally believe that agglutinative languages operate with postposition, while flectating
ones apply prepositions. While this
is true in many cases, I do not think we should adhere to this idea. Anyways,
Latin has a few postpositions (e.g.: Exempli gratia), and
Oscan likes postpositions [18]. However it seems to be easier to show the cases
and other relations at the end.
In present lots of
agglutinative languages are known. All Uralic languages are agglutinative. At
the Eastern end of the Old World Japanese is agglutinative and on the Western
end Basque.
The grammatical role
of a noun can be shown by the ending ("case") or by independent
words: prepositions in English or postpositions in Magyar. Japanese uses
exclusively postpositions, but there is no great difference between an
unchanged postposition following separately the noun and unchanged endings
glued to the root unchanged. In Magyar the difference is clearer; but that will
be Chap. 3 and the Appendix.
Different
agglutinative languages have different numbers of cases. As I told, that is 1
for Japanese. It is 5 (or 6) for Manyshi, nearest
relative of Magyar, and cca. 12 for
Finnish. That is complicated, ne ce pas?
3.
THE MAGYAR AGGLUTINATIVE CASES
My first language is
Magyar, the majority language of
There is no definite statement in the literature
about the exact number of Magyar cases. However the number is cca. 30. I have found a relative old university textbook [19]
trying to list them, and I start with this. However anybody can come up with
1-2 more cases, including me.
Following the Finnish
examples, Magyar cases have Latin names. But outside of
English equivalents
are given by prepositions. For Acc there is no preposition (English
prepositions go with Acc, so "on me", not "on I"), and
sometimes the preposition would not be unique.
Being Magyar a living
language, gluing together does imply minor changes. They are of 3 types:
1) If the ending
starts with a -v-, then this -v- changes into the end-consonant of the root, if
there is an end-consonant.
2) If the root ends with
a vowel, that vowel may lengthen; the rules are almost unequivocal, and if not,
both ways are good.
3) If the root ends
with a consonant and the ending is a consonant, then sometimes a vowel is put
in between; the rules are not always unequivocal, but often lax and never
important except style.
Magyar has a Vowel Harmony Rule, as Turkish and
Finnish do. So some endings have 1 form, some 2, some
3. The system is unequivocal, and will be shown in the Appendix. Here I give
the forms for words of back vowels, especially for doboz=box. The definite article ("the") is "a" before
consonant and "az" before vowel, but I omit
the article.
Direction Triality
is a general IE idea: for pronouns English has: where - whither - whence. Here
we start with nouns, but the idea of direction triality
is easiest to demonstrate with these 3 pronouns in English. In Magyar
adjectives and numerals do not differ for cases from nouns; pronouns are
slightly more complicated. In Magyar Nominative is the Absolute Case: no
ending.
|
Num |
Case |
Suffix |
Suffix classes |
Example |
English |
|
1 |
Nom |
- |
- |
doboz |
box |
|
2 |
Acc |
-t |
1 |
dobozt1 |
box |
|
3 |
Gen1 |
-nak a2 |
2 |
doboznak a |
of box |
|
4 |
Gen2 |
-é3 |
1 |
dobozé |
of box |
|
5 |
Dat |
-nak4 |
2 |
doboznak |
to box |
|
6 |
Loc1 |
-n |
1 |
dobozon |
on box5 |
|
7 |
Subl |
-ra |
2 |
dobozra |
onto box6 |
|
8 |
Delat |
-ról |
2 |
dobozról |
from box |
|
9 |
Iness |
-ban |
2 |
dobozban |
in box |
|
10 |
Illat |
-ba |
2 |
dobozba |
into box |
|
11 |
Elat |
-ból |
2 |
dobozból |
from box |
|
12 |
Adess |
-nál |
2 |
doboznál |
at box |
|
13 |
Allat |
-hoz |
3 |
doboizhoz |
to box |
|
14 |
Ablat |
-tól |
2 |
doboztól |
from box |
|
15 |
Fam1 |
-nott |
3 |
(Doboznott) |
at Boxes |
|
16 |
Fam2 |
-ni |
1’ |
(Dobozni) |
to Boxes |
|
17 |
Fam3 |
-nól |
2 |
(Doboznól) |
from Boxes |
|
18 |
Term |
-ig |
1 |
dobozig |
until box |
|
19 |
Lat |
-vá |
2 |
dobozzá |
turning into box |
|
20 |
Loc2 |
-t(t) |
1 |
dobozott |
at/on box7 |
|
21 |
Instr |
-val |
2 |
dobozzal |
with box |
|
22 |
Comit |
-stul |
2 |
dobozostul |
together with box |
|
23 |
Distr |
-nként |
1 |
dobozonként |
box by box |
|
24 |
Num |
-szor |
3 |
dobozszor |
as times as box8 |
|
25 |
Temp |
-kor |
1 |
dobozkor9 |
at box |
|
26 |
Caus |
-ért |
1 |
dobozért |
for box10 |
|
27 |
Quant |
-nyi |
1’ |
doboznyi |
as much/many as box |
|
28 |
Mod1 |
-ul |
2 |
dobozul |
as box |
|
29 |
Mod2 |
-lag |
2 |
dobozlag |
as box11 |
|
30 |
Mod3 |
-st |
1 |
dobozost |
as box11 |
|
31 |
Mod4 |
-lan |
2 |
dobozlan |
as box11,12 |
|
32 |
Mod5 |
-int |
1’ |
dobozint |
as box13 |
|
33 |
Mod6 |
-ként |
1 |
dobozként |
as box |
|
34 |
Mod7 |
-képpen |
1 |
dobozképpen |
as box |
|
35 |
Delat2 |
-ól |
2 |
dobozól |
from box14 |
1
"Direct object" (vs. subject, the Nominative).
2 The language frequently
avoids this case.
3 This is formally another
type of ending.
4 The ending is the same in
Basque. Perhaps the coincidence with IE is not accidental, but this is common
Palaeolithic inheritance.
5 The "indirect
object" of English.
6 The ending is the
same in Basque!
7 Rare.
8 As many times as...
9 Slightly forced with
"box"; e.g. "when reaches the box".
10 Both cause and consequence; "why?"
11 Answer to "how?"
12 Archaic.
13 With this noun it is rather unnatural.
14 Fossile.
For Column 4 see Appendix.
Table 2: The
Magyar nominal endings.
Sure, a few on
agglutinative "("case") endings are archaic, for which 31 &
35 are the best examples. 20 is also archaic, except
for geographic names, postpositions and pronouns; it was the original Locative.
I do not try to explain the differences of the modal cases 28-34 in English; in
Magyar it is clear. 35 is fossile
but it was/is the Old Delative and you find it in
many postpositions even now.
We have 4 (!)
direction trialities. The fourth one can be used only
with names (mainly family names), and only at some parts of the country. But otherwise...
Of course, not all of these
35 nominal cases are "absolutely necessary for communication", but
surely they do make the communication versatile. E.g. see Case 2, the
Accusative. Lots of languages do not differentiate between Nom and Acc; and
then the sentences "Peter beats Paul" and "Paul beats
Peter" differ only in word order; so the word order cannot be used then to
express something else, say emphasis. On the other way, proto-IE, older IE
languages and some modern ones do distinguish; see Latin: "Petrus caedit Paulum"~"Paulum
caedit Petrus"~"Petrus
Paulum caedit" = “Peter beats Paul” versus "Paulus caedit Petrum" = “Paul beats Peter”. Similarly, you do not have to distinguish for form between Accusative and Dative; and in many cases even with
wrong word order it is clear who gives whom to whom. However, imagine a simple
Latin sentence about slave market with me, Aulus and
Priscilla. The English scheme is: "I give A to B"="I give B
A". Now, "Do Aulum Priscillae" is clearly not "Do Aulo Priscillam", while neither is
impossible.
The 4 different
Directional Trialities clearly differ from each
other. 6-8 is about the outer surface
of "the box", and generally, but not always, the upper surface. 9-11
is about the inside, and 12-14 is
about the immediate neighbourhood.
While the fourth triality, 15-17 could be used only
allegorically about a box, it is quite regular for a family with the last name
Box. Then Doboznott means (I was) "at (members
of) the family Box", Dobozni is (I went)
"to (members of) the family Box" and Doboznól
is (I come) "from (members of) family Box". True, Budapest Magyar do not use this 3 forms; instead there it is told: Dobozéknál/Dobozékhoz/Dobozéktól,
an -ék- for "the group" and then Triality 12-14.
Terminative 18 can be
used either in space or in time; and it is not necessarily the final
place/time. Either it is, so the motion went until Point T and has stopped, or
we are still in motion but just now we have arrived at T so we moved just
"until T". Lative 19 is "to change
into" and such, in the sense of Mrs. Lot, who changed into a statue of
salt, so she “sóbálvánnyá változott”, where „sóbálvány” is the salt statue
and „változik” is „change into”. Second Locative 20 deserves a short Chapter of its
own. Instrumental 21 and Comitative 22 do often
coincide in 21 "when speaking loosely", as it does in English as
well, but an instrument is clearly not a partner. Distributive 23 has nothing
to do with Numeral 24: instead of "box" let us take "négy"="four". "Négyenként"
is used if I give four something to each individual, or I admit always 4
persons through the door, while "négyszer"
is used as "4 times..." either as a number or if a repeated act is
repeated 4 times or such; and they have nothing to do with Quantitative 27,
which is a quantity, "doboznyi pirula" is e.g. pills
which roughly (could) fill a box. Temporal 25 is to express a time: e.g. "Négykor" is more formally "at
The modal endings
28-34 are not always easy to translate into English without lengthy
circumscriptions, and "box" is not the best example for modal
finesses. But "angolul" is "in English" (speaking, writing,
&c.). "Tevőleg részt venni"
(where "tevő" is "doing" and
"részt venni" is
"to participate") is "to participate actively". "Örömest" is "gladly",
where "öröm" is "gladness"
("örömest"
is almost "örömmel", but not exactly).
"Ideiglen"
may sound archaic, but "ideiglenes" is commonly used as
"temporary". Case 32 is rare, but "szerint" is a very frequent
word meaning "according to"; "szer"
is "implement". Cases 33 & 34 are almost synonyms, but only
almost. "Kép" is "picture", and
"-képpen"
may mean that somebody acts "as if", or it may mean "in such a
way". But "úriemberként veszteni" is "to lose as a gentleman", and here I
could not write "úriemberképpen".
Case 35 is Delat2. Its
meaning is clearly a Delative and it occurs even if
it is fossilized. I included it to understand structures of some Postpositions.
Maybe Case 32 is dying; no new constructions are made. And Case 20 is special.
But otherwise the system is alive and becomes richer and richer.
4.
NA SLOVENSKE
Locatives of
geographic places (e.g. villages, towns &c.) are sometimes special. E.g. in
Russian, where Locative is preserved only as a case after some prepositions,
Prepositions "v" and "na" can
take Acc or Loc. With Acc they denote our Cases 10 & 7, "into"
and "onto", Illative & Sublative, while
with Locative case they are our cases 9 & 6, "in" and
"on", Inessive & First Locative.
However, a lot of Russian words, mainly geographic ones, have one Locative case
after not "v" or "na", and
another Locative case for "v" or "na".
(Really then the first one is not a Locative case at all, and indeed, some
Russian grammars call the case Prepositional, as if prepositions could not go
with Acc, Gen, Dat and Instr
as well in Russian.) E.g. take the Ukrainian peninsula
Magyar has an
interesting “anomaly”. Hungarian geographic names are generally in First
Locative, Sublative or Delative
when you mean "in, into, from", but places outside of
The title of this
Chapter is a Slovakian chauvinist slogan: "(You should speak) Slovakian in
Namely, in Slovakian
"in
Now, the situation is
even more complicated, but only with city names. Generally we use Loc1 with
Magyar city names, not Inessive; but not always.
There are no clear thumb rules for the
exceptions. Most geographic names of the home country come in Loc1, but some in
Inessive; and some can be also in Loc2 as well. If
the Loc2 form is possible, then generally there is an alternative, either Loc1
or Iness, but never both.
This Loc2 is old, not
very active, and can act outside of geographic names as well. E.g. "helyütt-helyütt" is "here and
there, but not everywhere" ("hely"="place").
"All together" is "együtt", where "egy" is
"1", so "all in one place".
5.
PRONOUNS
Pronouns are sometimes
more conservative than nouns. E.g. in English only some pronouns have Acc
forms: I/me, thou/thee, he/him, she/her, they/them and who/whom. Let us see now
the cases of personal and demonstrative pronouns. "Én"
is the Sg1 "I", and "ami, mi" is
"what". I give only the first 11 cases for economy. The demonstrative
pronoun is rather "ami" in statements and
"mi" in questions.
|
Num |
Case |
Sg1 |
Dem. |
English |
|
1 |
Nom |
én |
mi |
I, what |
|
2 |
Acc |
engem |
mit |
me, what |
|
3 |
Gen1 |
nekem a |
minek a |
of me/what |
|
4 |
Gen2 |
enyém |
mié |
my (mine), of what |
|
5 |
Dat |
nekem |
minek |
to me/what |
|
6 |
Loc1 |
rajtam |
min |
on me/what |
|
7 |
Subl |
rám |
mire |
onto me/what |
|
8 |
Delat |
rólam |
miről |
from me/what |
|
9 |
Iness |
bennem |
miben |
in me/what |
|
10 |
Illat |
belém |
mibe |
into me/what |
|
11 |
Elat |
belőlem |
miből |
from me/what |
Table 3: Some
cases of 2 pronouns.
Now, you can see that the demonstrative pronoun exactly
follows the nominal paradigm. But personal pronouns differ. In Accusative Sg1
and Sg2 does not take the -t ending,
but expresses Acc otherwise; and also in Gen2; and in the other Cases something
very similar to the nominal endings but somewhat longer come first, and then am -m to express Sg1 (and -d
for Sg2). The end-m/d is a regular feature in Uralic
languages to express possession, and you will understand everything after the
next Chapter.
Now, there are
demonstrative pronouns for spatial position, with primitive Direction Triality. Let us see mirror translations for an
interrogative one:
|
Where |
hol |
|
Whither |
hová |
|
Whence |
honnan |
Table 4: The
primitive Pronominal Direction Triality.
OK, the second is the Lative, but you cannot
recognise the endings in the first and third. They are old endings forgotten in
the case paradigm. The system has been rearranged in the next 2 millenia.
Now see a
Demonstrative Direction Triality:
|
Here |
itt |
|
Hither |
ide |
|
Hence |
innen |
Table 5: The
primitive Directional Triality of “here”.
We see the Loc2 in the
first line, a somewhat changed Lat in the second, and again the
unnamed/unnumbered case in the third. One feels that it is a Delative, but neither 8, the New, nor 35, the Old. OK,
pronouns can be more fossilized than nouns.
6.
POSTPOSITIONS
The case system contains
only 35 cases, so cannot be sufficiently versatile. To help it, there are
postpositions. E.g. we have seen that a "motion" in connection of the
surface of a box is expressed via First Locative, Sublative
or Delative, according to the Direction Triality. This is generally enough if the upper surface of the box is meant. But
if the lower surface is in context, then we take the old word "*al
> alj = bottom", and put proper (albeit old)
endings according to the Triality, and these upgraded
words we put behind the word for box. If my explanation is not clear enough,
the Table may help. There is an alternative: a postposition or a possessive
structure with Case endings on the possessed. And the two will not be
equivalent for meaning, although close enough.
|
Postposition construct |
Possessive construct |
English |
|
doboz alatt |
doboz alján |
under box |
|
doboz alá |
doboz aljára |
to below box |
|
doboz alól |
doboz aljáról |
from under box |
Table 6: A postpositional
Directional Triality
In the first line you see the proper Loc2 of the old word *al>alj=bottom in Column 1. This is used if the place is under the bottom of the box. However if
the place is exactly the lower
surface of the box, then it is "on the bottom of the box". The
"bottom of box" is "doboz alja", a correct possessive construct with modern
words in it, and then we append it with simply Case 6, the Loc1. When there are
proper cases, one uses the cases. Still, anybody can see the direct connection
between "doboz alatt" and "doboz alján" even if the first one uses Loc2, not Loc1.
The situation is
similar in Row 2. Al-vá>alá (Lat) ~ aljra (Sublat); surely Lative preceeded Sublative in the same way as Second Locative preceeded First Locative. (The extra -á-in "aljára" is for the possessive expression: "onto its bottom".) In the third row
"alól"
is an older variant of "aljról"; surely the -ól acted 2 millenia
ago and it is the Old Delative. (You may get the
impression that –ról
< -ra +
-ól, so Delat < Loc1 + Old Delat.)
This is not bad as first guess, even if the story was somewhat more
complicated.)
A very big variety of
postpositions exist and I doubt if anybody has counted them. "Fölött/fölé/fölül(ről)" is the same Direction Triality
with "föl" instead of "al":
"föl" is a quite standard word, now for the
uppermost layer of boiled milk, so these postpositions are used if we are upward
from the box. (In this work I ignore verbs. But I note that fel-/föl-
is also a verbal prefix meaning
"up".) If we are behind, we
take "meg/mög". Now the word denotes the
completion of acts, but several centuries ago it meant the farther side, and a
motion is more
or less completed if you have gone behind the barrier. So: "mögött/mögé/mögül": {Loc2/Lat/Delat2} of “mög”.
If the place is in front of the box, then you take "*el".
The front part of something is the "eleje"
of that something, so this is again the archaic form of the word, and then:
"előtt/elé/elől". If the place is besides the box, then we take the word "mell".
That now means "breast", but maybe two millenia
ago it meant "shoulder" too. For any case, if somebody stands besides me, then we stand shoulder to shoulder. So the right &
left sides of the box define places on the two sides "mellett/mellé/mellől".
Unfortunately there is no postposition (yet) distinguishing "right
side/left side".
Now
a brief excursion to history/genealogy. Our next linguistic relatives
are the Manyshi & Khanty
in Westernmost Siberia, so we can check the reconstructions on Manyshi. Most of our cases are postpositions there, so
maybe they were such in Old Magyar too (later). Our "mögött/mögé/mögül" is
"sist/sisigh/sisnöl" in Manyshi, and
"sis" is "backside". Now for the 3 endings, "igh" is
homologous with our Lative ending (-vá Case 19), -t is just our Second Locative (Case 20)
and -nöl
is our Third Familiar Direction (-nól, Case 17), and the end of it is homologous to our
"-ól",
35 Delat2, the Old Delative. So "mögött/mögé/mögül" is homologous
to "sist/sisigh/sisnöl", only the roots differ.
In some Manyshi postpositions the Magyar Lative
(Case 19) is substituted with the regular modern Manyshi
Lative. (I know the sentence looks strange. Evolution
is strange.) The modern Manyshi Direction Triality endings are: Loc/Lat/Abl:
-t/-n/-nöl.
(So their Locative is our Second Locative, their New Locative is our First
Locative and their Ablative is our Old Delative or
rather the Third Familiar.) Now an interesting Manyshi
Triality of Postpositions is the Modern and Quite
Regular "kiwört/kivörn/kiwörnöl", with the same meaning as our 3 case endings Inessive/Illative/Elative "-ban/-ba/-ból". No surprise: the root
is "kiwör"="the inside".
Now it is time to
return to the personal pronouns. We have seen some cases of "I"; now
I give 6 cases of "I" and "thou":
|
Num |
Case |
Sg1 |
Sg2 |
Sg1 |
Sg2 |
|
1 |
Nom |
én |
te |
I |
thou |
|
2 |
Acc |
engem |
téged |
me |
thee |
|
5 |
Dat |
nekem |
neked |
to me |
to thee |
|
9 |
Iness |
bennem |
benned |
in me |
in thee |
|
10 |
Illat |
belém |
beléd |
into me |
into thee |
|
11 |
Elat |
belőlem |
belőled |
from me |
from thee |
Table 7: Some
pronominal cases (of personal pronouns).
Now, the character code here is as follows.
Bold is the Sg1/2 possessive ending, while Italics is an ending triality of some old Case endings -n/-(v)é/-ől which
are Loc1/Lat/Old Delat. Now, what remained (except a
vowel -e- which is an old story and we can tell that it is for easier
pronunciation even if this is not true)?
We see that Acc comes
from Nom (and one more, Gen2, and no more). The Dative of Magyar is a mystery
up to now; but surely there was an old word meaning a goal of a motion. This (unknown) word was longer, but we can tell
for the moment that it was cca. "neke"~"being goal".
(In Manyshi our Case 5 is a Postposition, "nopöl", but its root is unknown.) Then "neke-m" is "my neke/being
goal" and "neke-d" is "thy neke/being goal". I do not believe that people reading
this site are specially interested in unsolved details
of the origin of Magyar Dative, so origin of "*neke" may remain temporarily hazy.
But the situation is
much clearer for Cases 9-11. There we see an old word, "*bele". Surely, the precursor of Iness/Illat/Elat
was a Direction Triality of Postpositions. (The
assumed l>n change in Inessive might come from
assimilation to the Locative -n.)
But what was the original root of the postposition?
In Modern Magyar
"bél" is "guts", but some
centuries ago it meant generally "insides". We had 4 kings called
"Béla", and of course that name did not
come either from dirty guts or from the feminine
Slovakian adjective "white". But the insides are priceless, well
defended &c.
Then
the original postposition triality cca. "belen/belvé/beleől"
with the root "bele"="insides" is
analogous with the Manyshi postpositional triality meaning our Cases 9-11 made from "kiwör"="insides".
And
now a check. The 3 endings appear in a prayer from 1195. Their forms
then were "-ben/-bele/-belevl", but the writing did not mark length, and the
probable reading is "-ben/-belé/-beleöl".
This is exactly a
Postposition Directional Triality formed from the
word "inside", only it is already written together with the previous
word. But in the prayer the ending does not yet follow the Vowel Harmony Rule
("milostben").
Postpositions never
take the vowel quality of the previous word. Well developed endings always do it (see Appendix). We do have records
that some endings were still postpositions some centuries ago, and Case endings
25 & 34 now still violate the Vowel Harmony Rule. So our present Cases 9-11
were a fresh ending triality in 1195, and earlier were
a postpositional triality as in Manyshi
now.
It is possible to
repeat this reconstruction with two other Case Trialities
(of course first/second personal pronouns cannot get the Familiar Triality). Let us see them in Sg1:
|
Num |
Case |
Magyar |
English |
|
1 |
Nom |
én |
I |
|
6 |
Loc1 |
rajtam |
on me |
|
7 |
Subl |
rám |
onto me |
|
8 |
Delat |
rólam |
from me |
|
12 |
Adess |
nálam |
at me |
|
13 |
Allat |
hozzám |
to me |
|
14 |
Ablat |
tőlem |
from me |
Table 8: The
Directional Trialities of nominal cases for a
personal pronoun.
Again, I marked the possessive ending; what
remains at the beginning, may be a postposition, changed in a millenium. Now, in Cases 6-8 the picture is clear: the
original root was something "*ra(j)" (I think nobody
is interested in tiny phonetic details) and then a Postposition triality ~"rajt/rajvá/rajól" was formed. In 1055
Case 7, the Sublative, is still a preposition, "rea". Its Loc member was unsuccessful to substitute
the Loc1 -n, but the Cases of the
personal pronouns are not the cases of a common root but mainly words with the
same personal possessive ending.
As for the Adessive/Allative/Ablative endings the above table shows
that originally they were 3 independent prepositions. I think, you are not too
interested, which ones.
7.
CONCLUSIONS
I think I demonstrated
that with agglutination even a system of 35 cases + more than a hundred
postpositions does not raise a problem. The rules are simple and the system is
not arbitrary. If a dictator made the population happy but in turn he wanted to
introduce 6 more Cases and 50 postpositions (partly for software operations,
partly for family structure of the dictator), people would need a year to learn
and then the bigger system would be as easy as the smaller one. If somebody
introduces an ending of, say, "-éstar/éster"
to refer to "transfer persons to the Second Reality”, then we can affix
this to any noun, even if they really cannot be transferred to the Net.
In Ref. [7] the Nantucketer expedition to Britain rescues a Princess of the
Kurlelo line (planners of Stonehenge, astronomers
&c.) whose first language is agglutinative and far relative to Tartessian, ancestor of Basque. The Nantucketers
are overwhelmed by the agglutinative language, so learn Proto-Celtic and import
a Tartessian captain to
I think I demonstrated
that with a Magyar in
And there was another
motivation behind this study as well. Remember, on All Saints' Day of Anno Domini 1612 in
APPENDIX: VOWEL HARMONY AND
NOMINAL ENDINGS
The phenomenon called
vowel harmony is present in the whole Turkic world and in Mongolian as well;
the rules there are rather similar than in Magyar. In the Uralic group the
vowel harmony system is full (but the system is slightly differs from the
Turkish one) in Finnish and rudimentary in Estonian. So I cannot prove that vowel harmony was present in
primordial Uralic, but it is rather probable because independent Turkic
influence on Finnish and Magyar is rather far-fetched. The Rule is very useful
for an agglutinative language, because it emphasizes the unity of the whole
word produced from a root with many further elements glued together.
The Rule in abstract
sense (valid both for Uralic and for Altaic languages) is the following. The
vowels are classified into groups A, B, ... The root
may contain only vowels of one group. Let it be A. Now, the postpositions have
alternatives PA, PB, ...,
differing only in the vowel(s), and a root of type A takes postpositions of
types PA; and so on.
The Rule was never
absolute. There are a few roots having mixed vowels. E.g. "apple" is
"elma" in Ankara Turkish; both Magyar and
Kazak correct it to the rule-abiding "
In Magyar the vowels
are classified into 3 groups according to Vowel harmony. But first we must list
the Magyar vowels; Magyar orthography is almost
phonetic, but not exactly.
First, Magyar vowels
can be short vs. long. This characteristics is indifferent
for vowel harmony. Length is indicated by prime(s) above the vowel according to
the suggestion of Jan Hus, famous heretics and
Bohemian nationalist leader at the beginning of XVth century, ending on the
stakes according to the decision of the successful Synod of Contance,
held under the auspice of the great King Sigismund I
of Hungary & Croatia, also Western Emperor and pretender for the Bohemian
throne. (Hus suggested such primes, hats, hooks
&c. also for consonants but they were not accepted by us and Poles.
Anyways, Hus was a heretic.)
Counting the
short/long variants too, there are 15 vowels, which may be Short/Long,
Back/Front, Open/Mid/Closed and Rounded/Unrounded,
using the English terminology. Magyar orthography (defined by the Hungarian Academy of Sciences) does not
distinguish between two short "e"-sounds, Open and Mid, in some
linguistic and musical publications written as e and ë, because the guy Kazinczy, having the biggest work in defining the new
Magyar literary norm at the beginning of 1800's, lived at a territory where e
and ë coincided. (This is just the birth area of Mihál
Kováč, second President of the Independent
Slovakia from 1993 too, so the ex-President speaks funny Slovakian as if he
were a Budapest Magyar, what he is not.) I will use here the Magyar linguistic
convention although it is not the best for foreigners.
As for objective
physical measurements, Magyar vowels are pronunciated
clearly even in unstressed position so the acoustic analysis is not a problem;
I refer to [21], [22] and citations therein.
Table 9 shows the qualitative properties of the Magyar vowels:
|
Orthography |
S/L |
B/F |
O/M/C |
R/U |
Closest familiar |
|
A |
S |
B |
O |
R |
Ĺ in Ĺngstrom |
|
Á |
L |
B |
O |
U |
Slovakian á |
|
E |
S |
F |
O |
U |
German ä |
|
Ë |
S |
F |
M |
U |
General e |
|
É |
L |
F |
M |
U |
Slovakian é |
|
Ö |
S |
F |
M |
R |
German ö |
|
Ő |
L |
F |
M |
R |
Long ö |
|
O |
S |
B |
M |
R |
General o |
|
Ó |
L |
B |
M |
R |
Long o |
|
I |
S |
F |
C |
U |
General i |
|
Í |
L |
F |
C |
U |
English ee |
|
Ü |
S |
F |
C |
R |
German ü |
|
Ű |
L |
F |
C |
R |
Long ü |
|
U |
S |
B |
C |
R |
General u |
|
Ú |
L |
B |
C |
R |
English oo |
Table 9: The system of
Magyar vowels
Now, the quality of
the root is classified as:
|
Root |
Property |
|
Back |
B |
|
FrontRounded |
FR |
|
FrontUnrounded |
FU |
|
Mixed |
- |
Table 10: The
Magyar roots and the Vowel Harmony classes.
The endings can belong
to groups as follows:
|
Alternatives |
As |
In Table 2 |
|
3 endings |
B/FR/FU |
3 |
|
2 endings |
B/F |
2 |
|
1 ending |
Indifferent |
1 |
|
1 ending |
Original B/F? |
1' |
Table 11: Vowel
Harmony and Magyar case endings.
Back roots generally
take Back Case endings if there are alternatives. Front roots generally take
front endings if there are alternatives; and for the trial endings they take
the one proper for the last syllable of the root. Unique endings of course go
with all roots.
As for the vowels, let
us see a list both for the root and for the endings (except for strictly unique
endings where there are no alternatives):
|
Vowel |
In root |
In endings |
|
A |
B |
B |
|
Á |
B |
B |
|
E |
F/FU |
F/FU |
|
Ë |
F/FU/- |
F/FU |
|
É |
F/FU/- |
F/FU |
|
Ö |
F/FR |
F/FR |
|
Ő |
F/FR |
F/FR |
|
O |
B |
B |
|
Ó |
B |
B |
|
I |
F/B |
1' |
|
Í |
F/B |
1' |
|
Ü |
F/FR |
F/FR |
|
Ű |
F/FR |
F/FR |
|
U |
B |
B |
|
Ú |
B |
B |
Table 12: The
system of Vowel Harmony.
This Table is not
complicated at all. For the seeming complications, ë in roots is sometimes
indifferent for B/F opposition; é is
the common long pair of both e and ë; and i/í are inheritors of 2
back/front pairs (present in modern Ankara Turkish) so endings with a single i/í may be really a B/F pair. In roots the orthography (and
maybe modern acoustics) does not show
the Harmony Class, but even unisyllabic "i" roots belong to definite B and F classes.
Now let us see one
example from Table 2 for every possible class shown above:
|
Type |
Case |
Endings |
Remark |
|
3S |
13 Allat |
-hoz/-hëz/-höz |
- |
|
2L |
08 Delat |
-ról/-ről |
@ |
|
2S |
05 Dat |
-nak/-nek |
- |
|
1' |
27 Quant |
-nyi/-nyi |
$ |
|
1 |
25 Temp |
-kor |
$ |
@ Long ó/ő in endings
evolved from two short vowels separated by some -gh-
or such. So there were no original 3L endings, and so no present *-ról/-ről/-rél triality, even
if [19] mentions a trial -tól/-től/-tél from XVIth century..
$ Type 1 endings are the
latest case endings, evolved from individual postpositions (which never take
the vowel quality of the
Table 13:
Examples of alternatives in endings according to the Vowel Harmony Rule.
Now a few words about
the physics: elementary speech acoustics. Sounds of speech are oscillation of
air, and the latter not being solid, simple longitudinal oscillations. So the
only degree of freedom is some real-valued function of time F(t).
Of course F(t) is directly measurable, but the fine structure of F(t)
is very probably unimportant. For first step we produce F(t)
as superposition of periodic terms (Fourier transform) as
F(t) = ∫A(ν)eiνtdν (1)
ν being the frequency. Then
f(ν) = A(ν)A*(ν) (2)
is the power
spectrum.
Now, for affricates,
Scottish r-type sounds or clicks f(ν) will not be
enough to characterize the sound of speech; but for stationary sounds will be (this is Ohm's Law), and, at least in
familiar languages vowels or sibilants are such. The power spectra are easy to
measure and for vowels they are smooth.
Half a century ago
Petersen & Barney demonstrated [23] that for American English the actual
vowel can be identified with some 90% success by measuring the frequencies of
the first two local maxima of f(ν); in the
acoustic jargon we tell : "the locations of the first two formants",
formants being the characteristics of the power spectrum f(ν), and a local
maximum is a characteristic part without doubt. It seems that the statement about the importance of the first 2 formants are
true for quite a few other languages as well. Further formants are either
unimportant, or they carry some other information (intonation, local timbre
&c.)
Now we investigated
this problem for Magyar vowels some 2 decades ago [21], [22] (and you may find
the citations therein useful); and here I recapitulate only one result.
Tables 9-12 gave the
system of Magyar vowels for formation and for the Vowel Harmony Rules. Now let
us see the first two formant frequencies; all short vowels except ë (which is
difficult to measure in
|
Vowel |
1st formant, Hz |
2nd formant, Hz |
|
Á |
757±10 |
1309±12 |
|
A |
554±8 |
986±14 |
|
E |
529±9 |
1713±21 |
|
Ë |
381±3 |
2162±33 |
|
Ö |
398±6 |
1516±22 |
|
O |
410±5 |
908±20 |
|
U |
292±7 |
811±10 |
|
Ü |
277±6 |
1812±28 |
|
I |
263±6 |
2262±30 |
Table 14: Magyar
male formant frequencies.
A pattern emerges. One
ultra-open vowel, á; then 2 open ones, a & e, with equal first formant frequencies, but for the
second ones higher for the front vowel. Then in the Mid
degree 3 vowels, ë, ö & o, with the same first formant frequencies but with
higher and higher second ones as B<FR<FU. Finally, at
the Closed degree 3 vowels, i, ü & u, the lowest
first formant frequencies (but again practically equal) and again the sequence
B<FR<FU for the second formant frequencies. (We were looking for
two i's, front and back, but were not yet
successful.)
We have the female
frequencies as well but maybe that would be too much. For any case, the first
formant frequencies are practically identical with the male ones; the seconds
are of course significantly higher.
So Magyar children
learn in early childhood from the mothers to produce identical first formant
frequencies for a & e, for o, ö & ë, and for
u, ü & i. Also, boys and girls learn to form
identical first formant frequencies
for the respective vowels in spite of the different sizes of the speech
channel; for the second formants they may fulfil the gender roles. Why? No such
thing in American English.
Who knows? But this is
surely a property of vowel recognition in the Magyar linguistic community. And
look: this is rather similar to Tables 11 & 13. An ending may not have
variants. But if it has, it may be 2, but then it is generally a/e or u/ü, or
may be 3, but then it is always o/ö/ë. So for the Magyar ears & brains the
vowels with the same first formant
frequencies are somewhat variants of each other: in the endings only the first
formant frequency is given, the second is determined and unequivocally
generated by the (second formant frequencies of the) vowels of the root. Is it
not interesting for a non-agglutinative, for an IE speaker?
The reader arrived at
this point (if at all) maybe asks: this is very interesting but why to read just this? Now, the Appendix
is given for two reasons.
First, it demonstrates
how automatic the system is, in spite of the multiplicity of endings because of
Vowel Harmony. I mentioned in Chap. 7 that even if a future dictator would
introduce by decree new words for his and only his blood relatives, people
could use immediately the words in any grammatical context. Now, Table 12 of
this Appendix shows how.
Second, I told that
the pronominal paradigm is slightly different from the nominal one; I gave
examples in Chap 4 Table 3. Now, I show that for personal pronouns the paradigm
is not only slightly different, but much wider: all postpositions are handled exactly as most Cases. Let us take
the Sg1 personal pronouns in 5 Cases and with 3 Postpositions. The Cases will
be Nominative, Accusative, and one of the 4 direction Trialities:
Adessive, Allative and
Ablative. Now, Postpositions are more versatile than Cases and are used if the
Cases do not offer fine enough distinction: e.g. the {Adess,
Allat, Ablat}
Direction Triality tells us only that the position
(static, end or start) is near to the
referred point. If you want to tell where
near, namely in front of, behind, above, below or besides, you must turn to the
Postpositions. If the position is near, beside,
then you use a Postpositional Triality {mellett, mellé, mellől}. So: doboz mellett = beside box. But if something is beside X, then it is at X, so doboznál, but not everything at the box is beside the box. However, without doubt, the Postpositional triality {mellett, mellé, mellől} is close kin
to {Adess, Allat, Ablat} for meaning. So this
Postpositional Direction Triality will represent the
Postpositional "Cases" of the Sg1 personal pronoun. The small part of
the comparison of the nominal and pronominal paradigms goes as follows:
|
Nom |
doboz |
box |
én |
I |
|
Acc |
dobozt |
box |
engem |
me |
|
Adess |
doboznál |
at box |
nálam |
at me |
|
Allat |
dobozhoz |
to box |
hozzám |
to me |
|
Ablat |
doboztól |
from box |
tőlem |
from me |
|
PP1 |
doboz mellett |
beside box |
mellettem |
beside me |
|
PP2 |
doboz mellé |
to beside box |
mellém |
beside me |
|
PP3 |
doboz mellől |
from beside box |
mellőlem |
beside me |
Table 15: Two
Directional Trialities for a personal pronoun.
Here italics stands for the
ending/postposition, while boldface
is the Sg1 "possessive" ending. As I told in Chap. 5, in most cases
of the personal pronouns "the endings come first and the person is
represented by the possessive ending; and in Chap. 6 I told that most Cases
evolved from postpositions, which in Magyar could, of course, be appended by
possessive pronouns.
However in (the
overwhelming majority of) the Nominal Cases the Vowel Harmony works, so now the Cases are indeed Cases, meaning
that they are shown by endings of integral words. But if the 35 cases of Table
2 form the nominal paradigm, then all the constructions similar to those e.g.
in Table 14 belong to the (personal) pronominal one. But then the latter is
longer than 100 rows; I think nobody ever counted them.
Now, it is not always
unequivocal to reconstruct the original postpositions which have degenerated to
case endings. However the cases of the personal pronouns still keep the
original class in Vowel Harmony, e.g.
in = -ban/ben, but
|
in me |
bennem |
|
in thee |
benned |
|
in him |
benne |
Table 16: The Inessives of the Magyar singular personal pronouns.
so surely the original postposition was Front, Unrounded and probably Open. Indeed, as already told, old
records suggest a form bel(e), whose close kin is bél =
guts, and belső = inside part.
I told that the origin
of the Adessive, Allative, Ablative Triality endings is not
so clear. But, anyways, the paradigm of the personal pronouns shows that the
origin of the dual Adessive ending -nál/-nél was Back (Open, and probably
originally Unrounded), most probably some *nagha; and that of the trial Allative
-hoz/-hez/-höz must have been Back, Rounded, Mid, and surely Short.
According to Berrár [19] the Khanty
has a postposition khosya
with exactly the same meaning as Magyar -hoz; the PFU *k+back
vowel evolves quite regularly into kh+back in Khanty and Manyshi but into h+back in Magyar, and in the Khanty
postposition the first vowel is indeed Back, Mid, Rounded and Short. As for the
dual Ablative -tól/-től, the same method +
educated guesswork would give Front Rounded, and the fact that there are not 3
representations of the Front suggests an original Close vowel and a second
syllable; something *tüghe → *tüwe + Delat2, the Old Delative.
Now, *tüwe is near to the reconstructed
old form of tő
= the base of the shoot, root. The slight change of meaning probably went as:
away from the root of sg → away from something,
but the details are not yet clear.
REFERENCES
[1] H. Everett:
[2] P. Anderson: Eutopia. In: H.
Turtledove & M. H. Greenberg (eds.): The Best Alternate History Stories in
the 20th Century.
[3] Poul Anderson: The House of Sorrows.
In: All One Universe. TOR,
[4] S. M. Stirling: The Sky People. TOR,
[5] S. M. Stirling: In the Courts of the Crimson Kings. TOR.
[6] E.
[7] S. M. Stirling:
[8] S. M. Stirling: Against the Tide of Years. Roc,
[9] S. M. Stirling: On the Oceans of Eternity. Roc,
[10] S. M. Stirling: Blood Wolf. In: H. Turtledove & Noreen Doyle
(eds.): The First Heroes. TOR,
[11] S. M. Stirling: Dies the
Fire. Roc,
[12] S. M. Stirling: The
Protector's War. Roc,
[13] S. M. Stirling: The
Meeting at
[14] S. M. Stirling: The
Sunrise Lands. Roc,
[15] S. M. Stirling: The
Scourge of God. In preparation, a few Chapters can be found on the Net
[16] Aristotle of Stageira: Opera omnia. Bekker N° 858a14-17
[17] E. Shapir:
Language: An Introduction to the Study of Speech. Harcourt, Brace & Co.,
[18] C. D. Buck: A Grammar of
Oscan and Umbrian. Evolution Publ.,
[19] Jolán
Berrár: Magyar történeti mondattan. Tankönyvkiadó,
[20] A. Favaro
& al.: Le opere di
Galileo Galilei; edizione nazionale sotto gli auspicii di sua
maesta il
re d'Italia. Barbera,
[21]
[22]
[23] G. E. Petersen & H.
L. Barney: Control Methods Used in a Study Of the
Vowels. J. Acoust. Soc. Amer. 24, 175 (1952)