Joual of Indian Philosophy (2006) 34: 87-139

DOl: 10.1007/sl0781-005-8189-0
FRITS STAL
Sp
r
inger 2006
PV11L1PL LPL1PLb
PLVLbb bL1Lb P L1N1L1ZP1Lb
anupisitavrddhiil vidyi nitiprasTdti
"science does not smile on those who neglect the ancients"
Bhartrhari, Vikyapadfya
INTRODUCTION
0u08Ih IDt8 0888V I08 8 DmO8ODDIC qu08IIOu! hOW 08u 8uPu8P0,
u8IuI8 OI 8IIIH0I8, 888I8I u8 Iu KuOW0dP0 O Ih0 WOId |WDI0h
Iu0ud08 OuI80V08)? 1 8h8 uOI lIV IO 0OuIIIDuI0 IOIh0 8u8W0I OIhI8
qu08IIOu DV m0I0 8D00u8IIOu, DuI 8I8II WtIh 8Om0 80I8. 8IIIH0I8
8uPu8P08 8I0 8I 088I 88 uDIV0I88 88 u8IuI8 8uPu8P08 8ud 8I0
u0IIB0II08III0I0d IOOu00IVIIZ8IIOu,VIZ., ID08O-080C08I0D WhI0h
1 08 uIO-PD0II08u, uOI 8I0 Ih0V 0OuHu0d IO m8Ih0m8II08, DhV8I08
8uU 8 0W OIB0I 80I0u008 IDI Iu uPI8h 8I0 OI0u I00II0d IO 88
`0X80I.
NV D8D0I 0Ou8I8I8 O OuI D8II8. 18II8 1 8ud 11 D8V0 Ih0 W8V OI
Ill 8ud 1N. 18II I d088 WIIh 8u0I0uI 8ud m0dI0V8 80I0u00. 1I8
ID08lI0 I8 P8I8, Ih0 8IP0 0OuIIu0uI IO WhI0D uIOD0 I8 8DD0ud0d.
8II 11 8K0I0h08 Ih0 80I0u008 O 8uPu8P0 ID8I d0V0OD0d Ih0I0 8ud
DIOVId08 8 D80KPIOuud OI Ih0 0m0IP0u00 O 8IIID0I8 8uPu8P08
Ih8I Wt D0-II08I0d Iu 18Il8 Ill, `8IV PIIID0I8 L8uPu8P08 8ud
1N, `1IOm 8IuI8 IO PIIID0I8 L8uPu8P0.
L0HuIIIOu8 0Om0 Iu Ih0 0ud, DuI Ih0 0Ou00DI O 8IIIH0I8 8uPu8P0
Ih8I I8 u80U h0I0 mu8I D0 08IIH0d 8I 088I IO 8Om0 0XI0uI. PIIIH0I8
8uPu8P08 0OuII88I WIIh u8IuI8 8uPu8P08 8u0h 88 uPI8h OI h8I.
Pu 8IIID0I8 8uPu8P0 I8 uOI u8IuI8OI8OII800m8. uI Ih0 ODDO8II0
dO08 uOI hOd. 8IuI8 8uPu8P08 m8V D0 8IIIH
0
I8 8ud 8IIID0I8IIV
8dmII8O d0PI008. 1uuPI8h, II I8u8IuI8 IO88V! the roof of the house.
O 88V! the roof the house has 800m8 IO m08u Ih0 88m0 DuI 8Om0
8D08K0I8 m8V 00 Ih8I II I8 8IPhlV 088 u8IuI8 OI mOI0 8IIIH0I8.
88 FRTS STA
bImI8IV, the chimney on the roof of the house I8 u8IuI8. uI *the
chimney the roof the house has has I88IIIH0I8 IuIh080u80ID8III I8 uOI
uPI8h |88 Ih0 88I0II8K Iudt08I08), 8ud V0I II I8 Iu 800OI08u00 WIIh
Ih0 88m0 Iu08 O 0mD0dUtuP IB8I 8I0 D8II O LuP

8h 8VuI8X.
P00OIdIuP IO L0OIP0 WI0I, 8 0W 18IV8Id uud0IPI8du8I08 0Oud Iu
1964 dOIWO8u0D0mD0ddIuP8, DuI0V0IVOu0h0dIIOuD0WIIhIhI00OI
mOI0. 180I8 IK0 Ih080 8hOW Ih8I 8Om0 O Ih0 Iu08 O 8u 8IIIH0I8
8uPu8P0 8I0 Ih0 88m0 88, OI 8I0 8ImI8I IO 8Om0 O Ih0 Iu08 O 8
u8luI8 8uPu8P0.
PIIIH0I8 8uPu8P08 8I0 uO! !h0 88m0 88 OIm8 8uPu8P08. 18 8
IOIm8 8uPu8P08W8V8 8IIH
0
I8 DuI 8u8IIIH0I8 8uPu8P0uOI8W8V8
OIm8 1 DuI ID080 qu08IIOu8 IO bO 00Im8u 8ud 88K0d Dtm IO
0XD8Iu IO m0 Ih0 dIH0I0u00 D0IW00u Ih0 !WO 8ud 08D00I8V Ih0
8IIIH0I8 V8II0IV. 1IO088OI00ID8u88Id, D88I08V, 1 KuOW 8 OIm8
8uPu8P0 Wh0u 1 800 Ou0 8ud 8dd00 8Om0IDIuP 8DOuI 8V8I0m8lI0,
0OmDO8IIIOu8l8ud 8VmDOI0. Lu8IIIH0I8 8uPu8P08 h0I00II00m0IO
IhI00W0D8II08Ih8I 8I0 t8I00 8IIh00udOIh0 V00I0u008 I8I. He dI0W
mV D8III0u8I 8lI0uIIOu IO Ih0 d0DuIIIOu Ih8I I8 PIV0u Iu lh0 IhIId O
Ih080 8ud IB8I I8 Cu0 IO VI0K 18IIt8Ou.
A artifcial language is a language that has been deliberately desiged by one person
or a small group of people over a relatively short perod of time. Synonyms for the
term artifcial language include planned language, constructed language, model
language, and invented language. Artcial languages designed for specific purposes
are also known by an array of other terms. Those used in works of fction are called
imagnary languages or fctonal languages. Those desiged to facilitate global
communications are called universal languages, auxiliary languages (auxlangs),
interlanguages or interlinguas, interational languages, etc. The realm of artifcial
languages also includes logcal languages, number languages, symbolic languages,
and pasimologes (gesture languages).
h08IIID0I88uPu8P08WIIh WhI0h W08h8 D00Ou00Iu0d IuIhI8 0888V
8I0 Iu0K0d 8W8V Iu Ih0 Hu8 DhI880 O IhI8 0uum0I8IIOu. h0V h8V0
D00u d08IPD0d OI u80 Iu 8
0
I0u008 8u0h 88 m8Ih0m8II08, OPI0,
uPUI8It08, DhV8I08 OI 0OmDul0I 80I0u00. h0V 8I0 OI0u OIm8. LV0u
0888I08 8uPu8P08 O 80I0u00 8u0h 88 L8IIu OI b8u8KIlI m8V D0
8IIIH0I8 OI OIm8 IO 8OD0 0XI0uI. NO8I O Ih080 8uPu8P08 m8V D0
8IudI0d OI Ih0tI OWu 88K0, WIIhOuI 0Ou00Iu OI lh0tI m00uIuP OI
80m8uII0 IuI0IDI0I8ItOu. h0V 8I0 IuV8I8DV 0OmDO8IIIOu8 OI
8VuI80II0 Iu Ih8I Ih0V DO88088 8VuI8X, 8 m00D8ut8U Ih8I 0u8D08 Ih0m
IO 0Ou8IIu0I 0OmDOuu0 0XDI088IOu8 IOm 8IDD0 Ou08 Iu 8 m8uu0I
Ih8I I8 8u8OPOu8 IO Ih0 0Ou8IIu0IIOu O 80uI0u008 |OI OIh0I
0OmDOuud 0XDI088IOu8) IOmWOId8IOIOID0I00m0uI8IV0XDI088IOu8)
Iu 8 u8IuI8 8uPu8P0.
ARTIFICIAL LANGUAGES 89
The most creative and powerful artifi cial languages are the
languages of mthematics. History shows that they have not been
'deliberately designed by one person or a small group of people over a
relatively short period of time.' On the contrary, these languages grew
very, very slowly. Some of their expressions took more than three
thousand years to arrive at their present forms which are, of course,
not their last. The modem symbol for negation, '-', for example,
started with a cuneiform expression in Old-Babylonian. The Indians
put a dot above a negative number, but the symmetries between
positive and negative and between addition and subtraction were
only expressed in a manner that made calculation wieldy when ' + '
and '-' came into use in the ffteenth century CE (Tropfe, 1980 : I,
144-146, 206). The development of expressions for equations took
equally long but was much more complex (Tropfe, 1980: I, 382 f.).
The language of mathematics is sometimes regarded as i it were
constant and invariable, a necessary reflection of the universe, a
metaphysical symbolism almost divine. An unnamed writer in The
Economist (December 20, 2003) declared that Latin was for Newton
'the closest approach in words to the utter directness of mathematical
symbols.' That does not only ignore history but puts it on its head. It
took almost a century before Euler, Daniel Bemouilli and others
translated Newton's laws from Latin into equations, making use of
Asian algebras that were streaming into Europe via the Arabs (Staal,
1995a: 75-76, 2004 and below). Leibniz went a step further and
declared that progress in mathematics is largely due to improved
notations (Mates, 1986, Ch. X), that is, to the development of
artificial expressions and languages.
I ANCIEN AND MEDIEVAL SCIENCE
1 . The Unity of Pre-Modern Science
'Science' is often described or imagined as 'Western', having
originated in ancient Greece with the Arabs acting as translators. It is
a prejudice of long standing and derives from an outdated picture of
the history of science. The history of pre-modem, that is: ancient and
medieval science can only be adequately understood i the Eurasian
continent is treated as an undivided unit. That insight evolved over
more than half a century, roughly speaking from Otto Neugebauer
and Joseph Needham to David Pingree, and is based upon the textual
and historical study of source materials in the classical languages of
90 FRTS STA
80I0u00 Ih8I Iu
0
ud0 PI8DI0, Ld-8DVOuI8u, LDIu080, LI00K, L8IIu
8Dd b8u8KIII.
h0800OmD8uVIuP 0h8II I1IPuI0) DOIII8V88Om0 O Ih0 DI8IOII08
I08IIOu8hID8 D0IW00u Ih080 0IVIIZ8lOu8. 1I I8 Ou0 O 8 8IP0 uumD0I
O DO88ID0 8ImDIH08IIOu8. V DI0IuI0 OmIl8 8uPu8P08 8ud 0uIII0
0IVIIZ8IIOu8, mIX08 P0OPI8DhI0 8DD08IOu8 WlIh 8uPu8P0 u8m08 8ud
0UDh88IZ08 88IIOuOmV 8u0 m8Ih0m8II08. V0u 8O, 8 0W IhIuP8 8I0
008I.08ODOI8mI8u8
0
I0u00I808II0I!h8u8uVOIh0I8ud h880X0II0d
8Om0 Iuu0u00, dII00IV OI Iu0II00IV, Ou LDIu8, LI0000 8ud 1uUI8.
LI00K 8ud 1udI8u P0Dm0ID08 800m IO D0 I08I0d IhOuPh III0 I8
KuOWuOOI8PI00duDOuhI8IOII080h8uu08.
J
h8II8ODVIOu8I8Ih8I
Ih0 PI8D8 8IOOd 8I Ih0 P0OPI8DD108 8ud ht8IOII08 00uI0I O
DI0-mOd0D 80I0u00.
LuOI8IIOum8IK8Iu\h00h8IIdI8W8Il0uIIOuIO8ImDIH08lIOu8ID8I
U8V D0 08D00I8V mI808dIuP. `100uI8lI0 I8 D800d D0IW00D quOI08
D008u80 0uP0D8u0I, WhO W88 0Ou00D0d IO 8hOW Ih8I uOI 8
100uI8II0 80I0u00 W88 LI00K, 0OuIIuu0d IO u80 Ih0 I0Im WhI0h
D0ID0Iu8I08 Ih0 VI0W Ih8I, Iu 80I, II W88 ~ 8u 0IIDu0Ou8 V0W I OuV
D008u808P0DI8, 8 DIVOI8-IODI0 Iu ou DI080uI0OuI0XI, dId uOI 0Om0
IOm lh0 8u0I0uILI00K8. LuOI8IIOu U8IK88uIIOuu0 `PI8D D008u80
Pt8D 80I0u00 dId uOI 8II80 IDm Ih0 PI8D D0
D
Iu8u8 IhOuPh II W88
P0u0I8V 0XDI0880d IhIOuPD Ih0 PI8DI0 8DPu8P0. `PI8D 80I0u00 I8
88OI00II0dIO88`188mI
0
,DuI8Iu00J0W88udLhII8II8u80OuIIIDuI0d
IO II, Ih8I 8DD08lIOu I8 0qu8V mI808dIuP.
h0 8D0 Ih8I 8Ouud8 08D00I8V DI0I0uIIOu8 I8 Moder. 1 h8V0 DOI
DuI II D0IW00u quOI08 D008u80 I D0I0V0 II I8 0OII00IV u80d. uI
modr I8 DOI Ih0 88m0 88 Wester OI Occidntal, I0Im8 Ih8I 08d 88
mu0h 88II8V 88 Ih0II 0OuuI0ID8II8 Easter OI Oriental ID8I 8I0 uOW
0Ou8Id0I0d Iu0OII00I. NOd0DZ8IIOu I8 Iuh0I0uI Iu 8uV OIm O
DIOPI088 DuI Ih0 I0Im I8 u80d Wh0u Ih0 0V0uI h881u8I O00uII0C. h0
moder O `mO00D80I0u00I00I8 IO \h0 mO8I I000uI 0V0uI O I8dI08
DIOPI088 Iu Ih0 I08m O KuOW0dP0. uI Iu OId0I IO D0 8D0 IO
d0I0ImIu0 hOW I8dI08 lI W88, W0 u00d lO 8dODI 8 WId0 D0I8D00lIV0.
1 I was wrong in claiming that the Vedo-Greek variety of geometry is diferent
from all Mesopotamian mathematics (Staal, 1999: 107) as Jens Heyrup has explained
to me. See Heyrup (1998: 32-40) which shows that Elements II.I-10 largely consists
of'cut -and-paste' results long known from Old-Babylonian sources. I like to use this
opportunity to clarify my statement, that Brahmagupta's expressions suggest the
notion of an equation (Staal, 1995a: 91). Suggestive they my be, but Hayashi (1995:
92, 333) has shown that they are not equations but tabular presentatons of the
numerical data from which solutions are obtained with the help of an algorithm.
ARTIFICIA LANGUAGES
A persistet but outdted picture
of the development of scice:
91
Gek

A ("tlatr")

us.-
A stil siplitic but more reaistic picture:
u., -=
+:::ac
" Id
+:a:
c ..
/::a:
a
"Ab"
cu..
> +:::ac
a:
!
Euope
> 150A
< u ) .
>
Figre 1. The Histor of Ancient and Medieval Science.
Modem sctence started with the scientific revolution which
occurred in certain sciences and in Europe though it is not fashion
able to refer to it as such. Many academics avoid the expression and it
is possible to ignore it when one's gaze is confned to Europe. If one
looks beyond that subcontinent, it is obvious that some such event
occurred there during that time. It explains that Joseph Needham,
whose work is not only about China but abounds in references to
Indian, Near-Easter and European sciences, was obsessed by the
question why the scientific revolution occurred in Europe and not in
China or India. Whatever it is, we shall see that the subsequent
explosion of mathematical activity through artifcial languages during
92 FRTS STA
te seventeenth and eighteenth centuries in Europe was perhaps the
greater revolution, and that is not the end of the story.
The arrows of Figure 1 do not only claim that the a:cient and
medieval science of the Eurasian continent preceded modern science
but led to it. That is an empirical hypothesis that we shall subject to
closer scrutiny in Section 4 'Rivers Flowing into the Sea: A Empir
ical Hypothesis'. Watever its empirical content, any such hypothesis
depends on non-empirical, a priori principles as well. One of these is
that an ancient problem cannot be understood unless subsequent
study has thrown more light upon the matter. Joseph Needham put his
finger on it when he wrote: 'to write the history of science we have to
take moder science as our yardstick-that is the only thing we can do
- but moder science will change, and the end is not yet' (Needham,
1976: xx). I refer to this yardstick principle as the principle of
modernity and since it comes first in terms of methodology, is widely
misunderstood and plays a central role in the following discussions, I
shall examine it in Section 3 'The Principle of Modernty' .
Needham's final clause - that science is never fi nal - applies to the
history of science itself. Assume that we were to discover that the
number system of the Mayans, which developed during the fst
millennium CE, derives from an ancient American civilization that
also inuenced Asia. Would it not afect our ideas about the early
history of mathematics? It is an unlikely possibility in the light of
what is presently known, but what will we know to-morrow?
Meso-American cultures contrast instructively with our 'Eurasian'
picture because they grew in total isolation until the sixteenth
century. No one has expressed it more clearly than Octavio Paz,
Mexico's great poet and her ambassador to India:
India was always in communication with other peoples and cultures of the Old
World: frst with Mesopotamia, and later with the Persians, Greeks, Kuchans,
Romans, Chinese, Afghans, Mongols. The thought, religons and art of India were
adopted by many Asian peoples; in tur, the Indians absorbed and transformed the
ideas and creations of other cultures. The Mexican peoples did not experience
anything like (that) ... They lived in an immense historical solitude; they never knew
the essential and common experience of the Old World: the presence of the Other, the
intrusions of strange civilizations with their gods, technical skills, visions of this
world and the next (Paz, 1995: 91).
Multiculturalism, thecorrectfashionespeciallyin the USA, treats Arabs,
Chinese, Euro-Americans and Indians as if they inhabited separate
cognitive worlds and were as isolated as were the Meso-American
cultures. That idea is not supported by the history of science.
ARTIFICIAL LANGUAGES 93
2. The Convergence of East and West: c 1200-1400 CE
Why do the Arabs stand at the geographical and historical center of
pre-modem science? Part of the answer is given by Figure 2
(Schwartzberg, 1978: 37) which takes us from history to geography
and from science to mlitary campaigns and conquests. The region in
the middle, surrounded by a thick black line, like the larger over
lapping area to its east, is the boundary of the empire of Timur or
Tamerlan around 1400 CE. It includes most of ancient Mesopotama
and a good part of the area of Hellenism but its center of gravity lies
further east. It was then and under Mongol rule that freedom of
travel and social intercourse were least restricted and hospitality to
strangers reached its highest peak (Schwartzberg, 1 978: 196-197).
The map of Figure 2 shows that anything like the so-called
scientific revolution could not have taken place before, say, 1 500 CE.
Prior to the twelfth century, many of the areas on the map were much
more advanced than Europe and we could, with hindsight, expect
them to have engendered a scientific revolution. Actually they did, as
we shall see, but in the sciences of language, areas to which most
historians of science do not pay much attention. In the mathematical
and physical sciences, the revolution happened later and further to
the west. By the twelfh and thirteenth century, 'Islamic and Christian
centuries were growing on roughly the same level and it is then
sometimes hard to know in which society a new development came
fst' (Hodgson, 1974, 11: 365). 1 conclude that it is not the place that
was dcisive but the time.
3. The Principle of Modernity
Like other sciences, the hstory of science itself does not only depend
on empirical facts but also on a priori principles. Conversely,
principles may depend on facts from the hstory of science. That
circularity or apparent
c
ircularity is in accordance with the absence of
principled boundaries between facts and theories that philosophers
like W.V.O. Quine and others have elucidated. No such circularity
implies that there are no facts at all and that the relativism that denies
their existence makes any sense. These problems are relevant to our
discussion but ths is not the plac to discuss them in any detail. 2
2 Moder philosophers of science do not often take account of the history of
science, let alone its non-European branches. la Hacking writes in the introduction
to his readings about scientifc revolutions: 'Newton's tr 'mass' may not even
Joph Sca
(e.)
A Hisorical Atlas of
South Asia
Chicago 1978
Figure 2. The Covergence of East and West: c. 1200--1400 CE.
-fl>ttl <f lff 1nol r.mpat llh
dtilt l CHlSlt k'hin, 111
fll.ltW <t M<I<! ttJ:ofr U:t
M fot"llt, 61 tht s, :; '
. itlf lt 4! tr1$ tl M)J)!
; :: ... ::::'i :;;:::g;
t:&o:
""' fta-tlotM.Fn!Cl, 'U1
"
-:
>
a
g
ARTFCIL LANGUAGES
95
I shall briefly return to relativism in connection with mathe
matics (in the present section) and to another principle that
is methodological or epistemological (in Section 5 'The Principles
of Acceptance'); but the principle of modernity that will be
considered here is diferent. Needham had a particularly clear
headed understanding of it and I have already quoted the way
in which he formulated it: 'to write the history of science we
have to take moder science as our yardstick - that is the only
thing we can do - but moder science will change, and the end is
not yet.'
Specialists in the history of Chinese scienc ofen criticize or even
ignore Needham. I cannot judge these discussions but some of the
critics seem to have already forgotten that Needham was a pioneer
who brought together 'the raw material on whch generations of later
scholarship can be founded' (de Solla Prce, 1971 : 1 7-18). Many
historians of science are similarly unhappy with regard to the yard
stick/modernity principle. I discussed some of these issues before
(Staal, 1993-1994: 9-19) but wilto take a fresh look at some recent
publications which are useful because they exemplify not what the
principle is, but what it is not.
G.E.R. Lloyd
G.E.R. Lloyd's recent comparison between ancient Greece and
China is leared and sophisticated and he is engaging writer. His
work is published in a distinguished Cambridge series, 'Ideas in
Context', which seeks to dissolve 'artificial distinctions between the
history of philosophy, of the various sciences, of society and pol
itics, and of literature'. Artificial things should be dissolved if it
can be done. It does not apply to artificial language itself despite
the belief of some so-called ordinary-language philosophers. But let
me not digress. The title of Lloyd's third chapter is 'The number
of things'. It begins with the following paragraph which throws
Footnote 2 continued.
mean what it does in Einstein's relatvistic physics' (Harkin, 1981: 3) Such insights
are obvious to historians of science who study, among oter things, changes in the
meaning of basic concepts, especially across language boundaries. As for relativism,
discussions continue unabated though it refutes itself: for any argument i support of
relativism must assume that the argument is relative and therefore not an argument.
See, e.g., Scharfstein, Staal and others in Ariel et al. (1998) and Scharfstein (2001).
96 FRTS STAL
light not only on our present discussion but on many topics that
will engage our attention later:
From the belief that mathematics reveals invariant truths it is all too tempting to
conclude that mathematics itself is invariant. That view is then ofen combined with
a thesis about the development of science, namely that it depends on a shift from the
qualitatve explanation of phenomena to their quantitatve understanding. That is
generally represented as the key factor in the changes that took place in what used to
be called the scentifc revolution of the sixteenth and seventeenth centuries, and in
support Galileo is regularly cited for the notion that the book of the unverse is
written in mathematical language (Llyod, 2002: 4).
Llyod continues: 'The faws in such a package of beliefs are obvious'.
The frst of these flaws, according to him, is writing history 'fom the
perspective of what was to come'. This may or may not be a veiled
attack on Needham but is diferent from the yardstick/moderty
principle. Llyod refers to earlier chapters of his book to support his
statement. What these chapters do, and what Llyod continues to do,
is stress the immense varieties we come across in history. He does that
with great verve and is fully convincing.
As I see it, there are four separate strands in the paragraph I have
quoted.
a. The frst is the relativistic stance that appears also elsewhere in
the book. It is particularly unpersuasive in the case of number and
mathematics in general. Mathematics does not seek to dissolve
'artifcial distinctions between the history of philosophy, of the
various sciences, of society and politics, and of literature'. It abstracts
from context, by defnition. It does not study 'two apples' or 'two
mangoes' but two. It is therefore by nature recalcitrant to context
which, in tu, does not throw any light on its nature. But does Lloyd
himself accept that mathematical truths are relative?
I don't believe that Lloyd believes that 2 x 3 = 6 is fawed, varies
over time or that we may accept with equal equanimity that 2 x 3 = 5.
If that were the case, he would experience continuous problems with
his fellow human beings, not to mention his calendar, plans or
recollections, and many other issues in his life. Without knowledge of
such properties of numbers, ancient Chinese and Greeks would have
had the same problems, not to mention great difculty in operating
their various types of counting-board. Nor is such invariance confied
to elementary arithmetic. '317 is a prime, not because we think so, or
because our minds are shaped in one way or another, but because it is
so' (Hardy, 1994, reprint: 130).
ARTIFICIAL LANGUAGES 97
Lloyd has a great way with words and is fond of looking at them.
He is inclined to restrict language to words without taking into
account that it consists of sentences and other compound structures.
In his chapter on 'The language of learing', he confines himself to
vocabulary and terminology, that is not merely words but mostly
nouns. In the chapter on number, he is concered to show that
Chinese shu covers a much wider range than Greek arithmos. It may
mean, for example, 'several', 'counting' and 'scolding'. Such uses
occur in many languages. Greek itself says fv &pt9J< Etvat 'being
counted' that is 'being honored', as well as fv ou oevi &pt9J< Elvct
'being despised'. English says: 'he made a large number of mistakes'.
Not only in mathematics itself, but even in a discussion of mathe
matics, the words of a natural language do not count for much - and
yes, I accept that that applies to my words, too.
{. Lloyd is admirably successful when he shows that the shift from
qualitative to quantitative explanation is an ancient one that occurred
in early China (as it did in India) and did not have to wait for Galileo.
It occurred in a variety of contexts and it is in the treatment of these
that Lloyd is at his best. And yet, later in the chapter, in the longest
sustained discussion of the contextual use of a mathematical figure,
viz, the gnomon (pp. 52-55), Lloyd himself frees himself from
contexts and his self-imposed relativism both. He concludes with
reference of calculations from gnomon shades, that the key diference
between Greece and China lies 'in the starting assumption made' and
adds that 'Chinese and Greeks methods were essentially the same,
mathematically'.
y. The next two strands present us with a puzzle. It is clear that the
flawed considerations of {. imply tht the shif from qualitative to
quantitative cannot be the key factor in the changes in 'what used to
be called the scientific revolution of the sixteenth and seventeenth
centuries'. We may accept the tiptoeing around that revolution as a
small bow to correct behavior, but does Lloyd deny that something
happened during that period? That would be odd, for although that
error is readily made by those who do not look beyond Europe, it is
an event that could not be missed by anyone familiar with develop
ments in other civilizations during that period. Lloyd need not be
obsessed by it, as was Needham, and it is true that Lloyd is primarily
interested in ancient texts. But why then does he venture here into
tose later centuries?
8. Deepens mystery because it involves Galileo who did write what
is always quoted in part or in ful, viz.: 'Philosophy is written in this
98 FRTS STA
grand book - I mean the universe - which stands continually open to
our gaze, but it cannot be understood unless one frst lears to
comprehend the language and interpret the characters which it is
written. It is written in the language of mathematics . . .' (quoted after
Blay, 1993/1998: 1). Lloyd has shown that this does not refer to the
transition from quantitative to qualitative. We shall see later what it
does refer to.
Elsewhere in his book, Lloyd pays due attention to the Euclidian
tradition of axiomatic-deductive demonstration which, in earlier
publications (Lloyd, 1990, 1996 a and b), he had discussed in relation to
Aristotle's and other philosophers' essays on demonstration as well as
the practice and theory of Greek rhetoricians and orators. He sees these
related uses as a proof that the Euclidian tradition is not merely
'intellectually attractive'. Here his relativist colors begin to shine more
brightly-but what, to a relativist, is a proof Lloyd confines himself to
two of the three levels that should be taken into account:
The frst level is that of Euclid. It is concered with mathematics.
It proves, for example, that the theorem of Pythagoras follows from
the axioms. If Lloyd does not accept that proof, 'intellectually
attractive' as it is, he should at least show that it does not so follow
which would ear him the Field Medal or Schock Prize, equivalents
to the Nobel.
The second level is that of physics. Euclid did not know physics -
or did he? We accept his parallel postulate because it describes a fat
surface. If we reject it, we can still use the rest of the system to
describe a curved one. Not just Euclid's system, but geometry and
mathematics in general, are known to apply to the universe, i.e., they
may be true. Lloyd is silent about it.
The third level is that of the derivative uses of such systems i
debates, rhetoric and oratory. Lloyd revels in it.
Are my first and second levels 'writing history from the perspective
of what was to come'? No, they show something else. Euclid's
demonstration is from axioms which, following Aristotle, he called
'common notions' (Kotvct Evvotct).
3
There was discussion about
these axioms (as Lloyd reports) and Euclid knew that his system
could be used with diferent axioms - axioms, after all, are
indemonstrable. It shows that Euclid understood that there could be
3 Euclidian deduction and his notions of geometrical construction are diferent
from Aristotle and modem concepts both. The same holds for the flourishing Greek
geometry in the third century BCE which is also independent of Euclid (aus, 1960;
Mueller, 1974, 1981, Harari unpublshed at the time of writing).
ARTIFICIAL LANGUAGES 99
'non-Euclidian' geometries. He may have been a schoolmaster rather
than a creative mathematician, but Euclid had a deeper under
standing of the efectiveness of mathematics in describing .the world
than Lloyd lets on. We shall return to this apparently mysterious
efectiveness.
David Pingree
Since examples abound, I move from Greece and China to India
where I am more at home.
David Pingree is the author of the monumental Census of the Exact
Sciences in Sanskrit and of numerous other publications that are
indispensable aids to all serious students of the subject. In a typically
solid but surprisingly technical contribution to a non-technical
magazine, Pingree sets the modern period aside in no uncertain terms:
One of the most signifcant things one leas from the study of the exact sciences as
practicd in a number of ancient and medieval societies is that, while science has
always traveled from one culture to another, each culture before the moder period
approached the sciences it recived in its own unique way and transformed them into
forms compatible wt its own modes of thought. Science is a product of culture, it is
not a single, unied entity. Therefore, a historian of premoder scientic texts -
whether they be written in Akkadian, Arabic, Chinese, Egyptian, Greek, Hebrew,
Latin, Persian, Sanskrt, or any other linguistic bearer of a distinct culture - must
avoid the temptation to conceive of these sciences as more or less clumsy attempts to
express moder scientific ideas. They must be understood and appreciated as what
their practitioners believed them to be. The historian is interested in the truthfulness
of the various sciences, not in the truth or falsehood of the science itself (Pingree,
2003: 45).
I see here again a variety of strands. The italics before the moder
period are mine because I like to underscore that the uniqueness of
the moder period in the manner in which Pingree singles it out has
nothing to do with the prnciple of moderty. The principle of
modernity is not a thesis; it is an a priori principle. It does not possess
any of the features that Pingree mentions.
I believe that Pingree would probably agree that his small Lloyd
like bow to fashion: 'science is a product of culture' does not mean the
same as 'science is not a single, unified entity'. I regard it as likely that
his statement is rather diferent from what some of the upholders of
political correctness have in mind. It is not science that is a product of
culture, but particular concepts of science along with particuar clas
sifications, evaluations and nomenclatures that are language-bound.
The last three sentences have me confused. Yes, we must avoid that
temptation about clumsy attempts that Pingree so clearly formulates.
100 FRTS STAL
But he does more than merely try to understand and appreciate these
ancient sciences 'as what their practitioners believe them to be'. He
adopts the Needhamian yardstick of moder science when he writes
about Indian mathematicians 'some were simply wrong' (p. 4 7) and
when he refers to 'accurate values', 'discovers', 'solving signifcant
problems' and uses phrases such as 'it had been realized in India' as
he repeatedly does (pp. 46, 50-52, etc.). Making such judgments is
not all there is to the principle of moderity, but it certainly is taking
moder science as our yardstick.
Johannes Bronkhorst
In his refections on Indian geometry, Bronkhorst (200la) adopts the
same yardstick as any reader can verify. He refers to some of the same
'mstakes' that were mentioned by Pingee whose assistance he
acknowledges. But he is especially concerned about the absence of
proofs.
shall not elaborate here on the special context in which these
investigations occur. In his 2001a article, Professor Bronkhorst has
paid careful attention to several of my eforts and essays. In addition,
he has invited me to participate in a conference on which more anon.
I am indebted to him and have leared much from his analysis which
has raised the discussion to a higher level. We agree on many basic
issues and one of these is the important position that grammar
occupies among the Indian sciences. But Bronkhorst's article is at the
same time part of a much wider agenda which consists of several
theses. He defnes human rationality in a special manner which he
sees exemplifed by the Greeks. He feels that Indians in general
possess a diferent kind of rationality and whenever they argue like
the Greeks - which is ofen - he wonders whether they might not
have been infuenced by the latter after the period of Alexander of
Macedon (who is no longer called 'the Great' by most Asians). That
was an important theme in the conference 'La rationalite en Asie'
organized by Bronkhorst in 1999 and edited in 200lb, and in several
other publications including Bronkhorst 200l c and 2002.
Both claims, about the absence of proofs or deductions and about
mstakes, have to be placed in the wider context of the history of
mathematics. There have always been two ways of doing mathe
matics. One is geometry; it emphasizes deduction. The other is
algebra; it uses computation and algorithms. The historian of science
who, to my knowledge, has most recently commented on some of the
ARTIFCIAL LANGUAGES 101
ramifcations of this grand division is Roddam Narasimha (2003 and
forthcoming and see note 1, above).
Lloyd (2002: 65) wrote about Euclid's axiomatic-deductive dem
onstration that it was common, though far from universal in Greek
geometry. Bronkhorst has an all-or-nothing attitude with respect to
the Euclidian system. Since he has not found in India, he has con
cluded that Indians don't have proofs. It is true that proofs are not
often given in texts, but that does not prove much in an oral culture
as Bronkhorst himself admits. It is also true that Indian geometers of
the Vedic period did not give axioms. But they provided deductions.
Several constructions are given by making use of the theorem of
Pythagoras or Baudhayana (an example is discussed by Seidenberg
1983 followed by van der Waerden, 1983; Staal, 1999). That is not
deduction from axioms but it is deduction as Seidenberg has rightly
emphasized.
There is another point about proofs and deductions that is relevant
to our discussion. The notion of proof, like that of exactness, is not
an invariant concept. Proofs possess a history during which standards
of rigor were considerably improved. In India, infite power series
were computed up to a certain point. They were recognized as
approximations that could be refned indefnitely and are infinite in
precisely that sense. But there was no proof of that recognition. In
Europe, the infinitesimal calculus was introduced and similarly used
for a long period without proper definitions or proofs. Greater pre
cision began to be cultivated in the nineteenth century and on the
European continent; the British Isles continued to lag behind.
Modern mathematics and logic present a more complex story. Few
people are competent to summarize it adequately and I am not one of
them. But it is clear that a Euclidian derivation from axioms has been
achieved only in some areas and has been proved to be impossible in
others.
Bronkhorst's second claim is that Indians made mistakes. They did
indeed and so did others. The list would be long so I shal confne
myself to Europe and mention two examples, one at beginning and
another at the end. About Plato, an eminent historian of logic has
written: 'the reading of his dialogues is almost intolerable to a logi
cian, so many elementary blunders are contained in them'
(Bochenski, 1951: 17; for Aristotle, see Patzig, 1969, Registr sub voce
and for Euclid: Daus 1960). My other example concers the proof of
Fermat's Last Theorem, which was finally announced, after some
350 years, by Andrew Wiles in Cambridge in 1993. Unfortunately it
102 FRTS STAL
0OuI8Iu0d 8 DI8I8K0 WhI0h II lOOK I08 8uOIh0I 0W V08I8 IO
0Imu8I0.
4. 'Rivers Flowing into the Sea . .. ': An Empirical Hypothesis
0 h8V0 IOu0h0U uDOD 80I8 DuI uOI uDOu Wh8I I8 D0Ih8D8 Ih0 mO8I
ImDOII8u! IhIuP 8DOuI Ih0m! IO D0 O u80 Ih0V h8V0 IO D0 0OuDIm0d
8ud Ih0I0 8I0 d0PI008 O 0Oum8IIOu. I008 Ih8I I W88 ID 18Ph8I
8ud 8W8V8 W8uI0d IO PO IO 18uI DuI 08uuOI I0m0mD0I 8uVIDIuP
d0DuII0 8DOuI II. 1 WOud 88V ID8I Ih0 OIm0I 80I, I II I8 8 00I, I8
DOI0 hIPhV 0Oum0d Ih8u Ih0 8II0I. 180I8 8I0 Iu Ih8I I08D00I IK0
Ih0OII08 8DOuI WhI0h VOP0I 10uIO80 (1997: 23 H.) I08 u8 Ih8I
L0u0I8 H08IIVlIV I8 0OuDIm0G IO Ou0 D8II Iu 10
14
8ud 0WIOu8
Ih0OIV OuV lO Ou0 D8II Iu 1 7. I 08uuOI 0OuDuI0 IK0 IhI8 DuI D0I0V0
lh8! m8uV P0DPI8DDI08 80l8 Ou Ih0 D8D O 1IPuI0 2 8I0 mu0h DOI0
hIPDV 0OuDIm0d lh8u DV mu8IuP8 8DOuI 18Ph8I 8ud 18uI. I 0Ou-
0ud0 Ih8I II I8 D0II0I I I8K 8DOuI empirical hypotheses Ih8u 8DOul
80I8.
00Ch8m h88 DIDVId0U u8 WIID 8 VI8IOu WhI0h, IK0 8 VI8IOu8,
8I8Dd8 Iu u00d O 0OII00IIOu. LuI 1IPuI0 1, WBI0h I8 D880d uDOu lD8I
VI8IOu, DI0IuI08 8u 0mDIII08 DVDOIh08I8 Ih8I u80uV 0OII00I8 WId0-
8DI08d 8ud D0I8I8I0uI 8IIIIud08 8ud DI0judI008. uI II I8 IOO 08IV lO
d008I0 ID8I II I8 hIPhV OI 0V0u WId0V 0OuHIm0C. Lu0 O ^00dh8m8
D0I8DDOI8 I8 D8IICu8IV DI808dIuP. 10 8Om0IIm08 D8IuI8Iu8 Ih8I
8 80I0u008 O Ih0 D88I HOW tuIO Ih0 80I0u00 O Ih0 uIuI0 IK0 IIV0I8
HOWIuP IuIO Ih0 808. 1I CId uOI 8W8V8D8DD0u D008u80 8Om0 80I0uIIH0
Iu8IPhl8 U0V0OD00 Iu I8D8IIOu, OID0I8 h8d uO ImD80I Ou 8uD800u0uI
d0V0ODm0ul8 8ud m8uV m8V h8V0 dI88DD08I0d WIIhOuI 08VIuP 8
II800D0hIud.B00880O m00I0Iu0I8 Ou0OIh08008808 D008u80Du0h
O II d0V0OD0d Iud0D0uU0uIV Iu Ih0 PI08I 0IVIIZ8IIOu8. LV0u 8O, Ih0
LDIu080 W0I0 880Iu8l0C DVudI8u m00I0Iu08ud 8m8 0uII0uI8 OIh0
PI08I 8II08m O 80uDuu0IuI0 8I0 D0PIuuIuP IO D0 8dmIII0U DV m0dt08
08I8DI8Dm0ul8 0V0u Iu Ih0 LuIO-^m0II088. 1I I8 IK0 Ih0 1udI8u
PVuIV0d8, uOW D8II O 8 DIIOIO O8I0Iu8IIV08ud 0OmD0m0uI8IV
Ih0I8DI08 O0I0d 8OuP8IC0 DOd0ID DIOm0dI0Iu0 luj88IVK, 2003:
407). 1I 00II8IuV t8 8 80I ID8I 80I0u00 h08 8W8V8 II8V00U IOm Ou0
0uIuI0 IO 8uOIh0I, 88 1IuPI00 DuI II Iu ID0 D8888P0 I quOI0d 8 0W
D8P08 08Il0I. 00dh8m 8ud 1IuPI00 h8V0 PIV0D Iuuum0I8D0 0X8m-
D08 O IhO80 II8V08 Iu uum0IOu8 DuDlI08IIOu8, 8 0W O Ih0m I8I0d
D0OW Iu References.
ARTIFICIAL LANGUAGES
103
h0 h0IO00uIII0 0880 I8 uu80IIIuP IO 00dh8m8 VI8IOu 8Hu 800I8
8uOIh0I hI8lOII8u O 80I0u00 WhO W88, Iu 8ddIIIOu, 8 0I08IIV0 m8Ih0-
m8II0I8u. PII8I8I0hO8 O b8mO8 d00ud0d Ih0 Id08 U8I Ih0 8uu
D00uDI08 Ih0 00DI0I O Ih08O8I 8V8l0m Iu 280 L1, A8D
]
8I8 Iu476
LL, 8Dd LOD0IuI0u8 Iu 1543 L. LOD0IuI0u8 Ku0W It8!8I0hu8! Ih0
m8Iu dI0I0u00 D0IW00u Ih0 IWO t8 uOI Ih8I Ih0V DOIuI0d OuI 8Om0-
IhIuP CI0I0uI, DuI Ih8I LOD0D0u8 W88 8u00088u Iu 0OuVIu0IuP
OIh0I8 8ud PII8I8I0hu8 W88 uOI. 0IIh0I W88 A8Dh8I8 IY 8uO,
1980). uI uOW 8 I0m8IK8D0 Iu0Ou8I8I0u0V 8BOW8 uD Iu ID0 WOIK O
.L. V8u d0I 80Id0u. 1I8 m0IhOdOOPV W88 IO 8W8V8 888um0 Ih8I
IhIuP8 8I0 I08I0d D008u80II m8V dI8D8VIuI0I08IIuPIuI0I0Ouu00!IOu8.
1u 1969, h0 III0d IO 8hOW IB8I

IV8Dh8!8 Ku0W PII8I8I0hu8, DuI uOI

ID8I LOD0IuI0u8 Ku0W A8Dh8I8. 10Ih8D8 lh0 8II0I hVDOIh08I8 I8
mOI0 8I-0I0h0d Ih8u ID0OIm0I, DuI WhV dId V8u d0I 80Id0u uOI
DuI8u0 II?
PuOIh0I 0OuuI0I-0X8mD0 IO 00dh8m D0II8Iu8 IO m8Ih0m8lI08
8ud I8K08 u8 0O80I IO OuI IB0m0 ~ 8IIIH0I8 8uPu8P08. 1I I8 DV uOW
D00OmIuP mOI0 WId0V KDOWu Ih8I IuDBII0 DOW0I 80II08 Ih8I 8I0
0XD8u8IOu8 O n 8ud IB0 IIIPOuOm0III0 uu0!IOu8 W0I0 dI80OV0I0d Iu
b0I88, 8OuIhW08I 1udI8, DV N8dh8V8 WhO HOuII8D0d duIIuP 1380/
1420 L1 (800, 0.P., 1IuPI00, 1981: 49-51, 65-66), 8mO8I IhI00 00u-
IuII08 D0OI0 Ih0V W0I0 dI80OV0I0d Iu LuIOD0 DV LI0POIV, 0W!Ou
8ud L0IDuIZ. uI Ih080 1udI8u UI80OV0II08 W0I0 OImu8I0d Iu 8
0OmD0X OIm O b8u8KIII 8ud dId uOI, 88 8I 88 I8 DI080uIV KuOWu,
HOW IuIO Ih0 80I0u00 O Ih0 uIuI0.

h0 88m0 m8Ih0m8II08 0X8mD0 0d lO dI0I0uI I08ul8 Iu LuIOD0

Wh0I0 \h0V 8uDDOII 00dB8m8 0mDIII08 BVDOIh08I8. 1u uIOD0,
IuHuII0 DOW0I 80II08 0d IO ID0 080uu8 8Hd W0I0 8 DOW0Iu IuPI0-
dI0uI O Ih0 8U0uIID0 I0VOuIIOu. 1I I8 IIu0 Ih8I 0WIOu8 L8IIu W88
88O OD80uI0, DuI LuIOD08u8 D0P8u IO 0XDI088 KuOW0dP0 WI!h Ih0
h0DD8P0DI8I0uDI8IIOu8Ih8IW0I0088VIO 08D8ud 8OOu8DI08d8
OV0I Ih0 WOId. 1I 0XD8Iu8 Ih0II I0VOuIIOD8IV ImD80I 8ud I8 Iu
800OId8u00 WIIhL0IDuIZVI0W, Ih8I DIOPI088Ium8lh0m8II08I88IP0V
Uu0 IO ImDIOV0d uDI8IIOu8.
4 Results of the Madhava school reached the Tamil country but no other places i
India where mathematcians were writing in Sanskrit and could have understood
them. There is cicumstantial, but no direct evidence to support the idea that some of
Madhava's discoveries may have reached Europe. A group of scholars at the Un
versity of Exeter is collecting more information (The Arabhata Group, 2000). One
difculty is that the Sanskrit of Madhava and his followers is much more difcult
than the Arabick that European scientists since the Renaissance were eager to lear.
J+ FRS STAL
L0IDuIZ DdV0d d K0V IO0 Iu Ih0 0u!II0 d0V0DDm0uI. DIOuPhOuI
DI8 I0, h0 DuI8u0d Ih0 Id0d O 0I0dlIuP d uDIV0I8d, DhIO8ODhI0d
duPu8P0 Ihd WOud D0 du Id0d duPudP0 OI Ih0 0XDI088IOu O
IhOuPhI WhI0h, OI hIm, Iu0ud0d WhdI I8 uOW OI0u I0I
=
II0d IO d8
80l0uIIH0 KuOW0dP0. O O!h0I mdID0mdII0Idu hd8 IuIIOdu00d d8
mduV 8VmDO8 IhdI dI08II Iuu80. LII00lV du0 IudII00IV Iu8DII0d DV
Ih0dP0DIdIhdI hdU D00DImDOII0d VIdIh0PIdD8, 0IDuIZ uOIdIIOu8
OW0G IuIO ID0 80d ODOd0Iu 80I0u00.
5. The Principle of Acceptance
LIdmmdI OI IuPuI8II08 I8 duOIh0I 80I0u00 lhdI D!8 00dhdm8 VI8IOu.
1IIdK08u8 lO mV0Dl0IODI0 WhI0hI8 IO8BOW, !hdIdud WhVOImd OI
dIIIH0Id duPudP08 O00uI d0ID88 80I0u008 dud 0IVIIZdlIOu8. 1I8 Iu0u-
8IOu h0I0I8d8OIud00OIddu00WIIhduOIh0ID
I
lu0ID0OIh0m8IOIVO
80I0u00. 1 I00I IO II d8 `h0 1
I
ID0ID0 O P000DI8D00! When studying
the history of seience at an earlier period or in another civilization, we
should not blindly follow our own principles, concepts and categories of
science, but accept those that were prevailing then and there.
h0 P000Dldu00 1IIu0ID0 I8 uOImuP u0W Iu Ih0 0DmDdIdIIV0 8IudV
O 0IVIIZdIIOu8. I!h I0PdId IO 1udId, OD0 O ID0 0880uIId 088Ou8
IduPhI DV 18uI0 1DPd8 Id8 OImudI0d DV 1OO0K, V3. +VV WIID
8D00Id I0I0I0u00 ID 1uP88, >DJ) W88 `hOW ImDOIIduI II I8 !O IdK0
d00OuDI O IIdUIIIOud 0dI0POII08 dud 0Ou00DI8 Wh0u dIl0mDIIuP IO
uud0I8IduC Ih0 0uIuId d0Dl0V0m0uI8 O du0I0uI 1udId. P 8ImD0
Iu8lIdIIOuOIh0P000DIdu001IIu0ID0I8Ih0I0Imscience Il80OIIII8
DV uO m08u8uuIV0I8d. 1l8Pu9O-Pm0II0duu800mDhd8IZ08WhdI dI0
8Om0IIm08 0d0d the exact sciences, du 0XDI088IDu P0u0IdlV 0DuHu0U
IOmdIh0md!I0d80I0D008dud80I0u008OudIuI0.1I0X0ud08Ih0hum8u
dud8O0Id80I0u008, dud 8Om0!Im08!h0I080I0u008d8W0, dOIh0m
dt80IDIu08IhdIdI0Iu0ud0dIu1I0D0hscience, L0Imdu Wisenschaft OI
1dDdu080 gaku. LDHI8h u8dP0 OIIPIudI0d Iu uIu0I00uIh 00uIuIV
L0Imdu DhIO8ODhV IDdIII0udIV Ih0m LIIh0V! bIdd VVU,VV,
LhdDl0IZV).1W0PODd0KuIIh0IIuIOLuIOD0dududPIdDDl8IOIVW0
Hud IhdI PIdmmdI, Ih0IOII0, O9I0, dud mu8I0 W0I0 I0PdIC0U d8
80I0uIID0 dI80IDIu08 dI 0d8I uDlI 0WIOu. Nu8I0 Iu0ud0d uOI OuV
d0Ou8II08 dud mu8I0d !h0OIV DuI d8O Ih0 dIl O 0OmDO8IIIOu I0.
LOh0u,V+). 18d0h8 Art of the Fugue uOI 0Xd0I dud80I0DIIH00V0u
IhOuPhII I8 d8OI0P8IC0dd8dII dud uOI 00dI WhdI II I8 d 80I0u00 O
h0 I0dd0I mdV WOud0I Wh0ID0I I 8m dIPI088IuP DuI lh0I0 I8 d
I0d8Ou. 1 Du8I 0Ou088 IhdI 1 dI0I OuIhI8 I88u0 WIIh LdVId1IuPI00.1
ARTIFICIAL LANGUAGES 05
8m uOI quIDDIuP h0I0, 881 dIUIu Ih0DI0VIOu8 800IIOu, WIIh I0P8Id IO
8Om0 DDI8808 Iu 8 8m8 8ud D0Ih8D8 Iu0Ou80qu0uII8 D8D0I, DuI
I8I8IuP 8 qu08IIOu Ih8I 0Om08 IO !h0 OI0 Iu Ih0 8uIhOIIl8!IV0 8u0
mOuum0uI8 Census of the Exact Sciences in Sanskrit. P8IuPI00I08
u8 DIm80 tu Ih0 Introduction IO Ih0 800Oud VOum0 I1V!), `Ih0
8uIhOI8 0Ou00DIIOu O Ih0 80OD0 O Ih0 WOIK h08 DIO8d0u0d. h0
0X80I 80I0u008 h8V0 uOW 0Om0 IO tu0ud0, 8OuP WIIh D8Ih0m8!I08
88IIOuOmV 8ud hOIO80ODI0 88IIOOPV, Ih0 80I0u00 O UIVu8lIOu,
P0OPI8DhV, 0O8mOOPV 8Dd OIh0I dI80tDIu08 ID8I IuVOV0 Ih0 d0I0I-
DID8IIOu O Ih0 DIOD0I IIm08 OI Ih0 D0IOIm8u00 O IIIu8 80I8. Ad
V0I, Ih0I0I8uOm0uIIOuO PI8Dm8I WDl0hI8ID1udI8Ih0 D8I8dIPmO
Wh8I 8u exact science 8hOud D0 IK0.
1IuPI008 0mDh88I8 Ou Ih0 0X80I 80I0u008 0Om08 IOm 0uP0D8u0I,
WhO WIOI0 mO8IV 8DOuI N08ODOI8mI8 88 uO08 1IuPI00 8DOuI uUI8.
10Ih8D8 0uP0D8u0I W88 IIPDI 8DOuI N08ODO!8mt8 8IhOuPh PI8m-
m8Il08 I0XI8 0XI8I0d Ih0I0 IOm 8IOuud 1600 L, Wh0u Ih0I0 W88
0Ou00Iu 8DOuI bum0II8u D00OmIuP OD8O0I0. I 8m uOI 0OmD0I0uI lO
D888 judR0uI Ou Wh0Ih0I Ih0 PI8mm8I Ih8I I8 GI80u880d Iu Ih080
I0XI8, 8ud 8DOuI WhI0h Ih0I0 0XI8I8 8u 0XI0u8IV0 lII0I8IuI0, 8DOud D0
I0P8Id0d 08 8u 0X80I 80I0u00. P8 OI 1udI8 8ud b8D8KDI, II h88 u0V0I
D00u dOuDI0d Ih8I PI8Dm8I W88 DI00I80V Ih8I, WhI0h UO08 uOI ImDV
Ih8I II W88 8I 8uV IIm0 D0I00I DI I00 IOm 0IIOI.
I 8m uOI 0OmD8IuIuP O ID0 80Iu8 ODt88IOu O Ih0 PI8mm8II08
80I0u008 IOm 1IuPI008 magnum opus OI WhI0h Ih0I0 8I0 POOd
DI80II08 I088Ou8. 1 PI8mm8II08 0OuIIIDuIIOu8 W0I0 Iu0ud0G, lh0
8IZ0 O Ih0 Census mIPhI h8V0 IO D0 mOI0 ID8u dOuD0d I800, 0.P.,
b0h8I0, V!!, 8 WOIK ID8I 8DD08I0d Iu Ih0 88m0 Du!I-VOum0
History of Indian Literature 88 1IuPI00, V). uI 8hOuduI Ih0II
0XI8I0u00 8I 008I D0mentioned? 18uI ID0 u0P00I O IuPuI8II08 8 OIm
O 0u!uI8 DI0JudI00 du0 IO Ih0 80I Ih8I LuIO-Pm0II08u 8Iud0uI8 O
8uPu8P0 0mD8IK0d OuV I000uIV I8ud uOI WIIhOuI 888I8I8u00 IOm
Ih0 1udI8H II8UIIIOu) Ou Ih0 80I0uIIH0 8IudV O Ih8I dI80IDIu07
1 W0 0X0ud0 Ih0 mO8I I000uI-D0IIOd, 0du08I0d 1udI8u8 h8V0 OI
mt0u
D
I8I0P8Id0d II 88ODVIOu8Ih8IPI8B8It8 uOI OuV`Ih080I0u00
O lh0 80I0u008 (sistrifll sistram), DuI 88O Ih0 mO8I 0X80I 80I0u00.
LuuI088 I0X8 8RIm II 8ud V0I80 000DI8I0 II8 IuHDtI0 0XI0u8IOu.
Boundless indeed is the science of languge,
But life is short and obstacles are numerous.
Hence take what is good and leave what is worthless
As geese take milk from the midst of water.
U FRITS STAAL
O uud0I8!8ud Ih0 mO8I DODu8I DIOV0ID Ih8I Iu8II8I08 Ih0 DI0VIIV
O PI8Dm8!I08 Iu08 Ih8I PI8mm8II8D8 800K, W0 mu8I XuOW 8 80I
8DOuI 1udI8u 0uIuI0. N8uV 1udI8u8 DI00I8Ou8 IOd8uPhICI8 D008u80
8Ou8 D0IOIm III08 Ih8I DIO088 IO I8K0 08I0 D Ih0II D8

0ul8 Iu Ih0
8I0II0. 1 Ih0I0800I ODIu08 Ih8I 8u0h 8 D0I0 I8 DOIII08V Iu0OII00I
8ud]OI8 8uD0I8IIIIOu, Ih8II8Hu0WtIhm0, DuI II0XD8Iu8 Ih088VIuP IO
WDI0h1 W8uI!OdI8W8!I0uIIOuD008u80II Iu8II8I08Ih00XII8OIdID8IV
0Ou00Iu OI 0X80IIIud0 Ih8I 0D8I80I0IIZ08 Ih0 1udI8u 80I0u00 O 8u-
Pu8P0! `PI8mm8II8u8 I0jOI00 OV0I Ih0 88VIuP O h8 8 8V8D0 Iu 8
PI8mm8II08 8I8I0D0DI 88 OV0I Ih0 DtIIh O 8 8Ou.5
0I0 18DIuI 8ud OIh0I 1udt8u PI8mm8II8u8 IuI0I08I0d Iu DI0VIIV
OIII8 OWu88K0?h0I0D88D00umu0h CI80u88IOuO!h8Iqu08IIOuDuI
D8I8KV IVV) h88d0mOu8II8I0d Ih8I 00OuOmV W88 8uIuI8 W8V O
80hI0VIuPP0u0I8IZ8IIOu. D8I8KV 8W8V8 8dh0I08 IOIh0 DIIu0ID0 O
mOd0IuIIV. 10 dO08 uOI II08I 18DIuI 88 8u 0XOII0 OD)00I O DDIO!OPV,
DuI 88 8 0O08Pu0 O P0uIu8 WhO d080IV08 IO D0 !8K0u 80IIOu8V 8ud
m8V D0 IIPhI O
r
WIOuP.
h0 ImDOII8D00 O PI8mm8I Iu 1udI8u 0IVItZ8IIOu h88 D00u
0XDI0880d DV 1OuI8 V0uOu Iu DI8 0I0IIOu8 m8uu0I! `PdhCI0I 8 !8
D0u8C0 IudI0uD0, 008I d8DOId D0u80I 0u PI8mm8III0u.
II THE SCIENCES OF LANGUAGE AND THEIR APPLICATIONS
6. Linguistics: An Apparent Tunnel
h8I Ih0 0DI0 0Ou00Iu O Ih0 1uCI8u 80I0u008 D88 8W8V8 D00u WIlh
Dum8u 8uPu8P0 I8 8DD8I0uI IOm ID0 08II08I 0888ID08IIOu8 O 8U-
0u008. uIIuP Ih0 8I0 N0dI0 D0IIOd, 8IOuud !h0 DIdd0 O Ih0 D8I
DI0uuIumLL, 8 I8I I8 PIV0u O 80I0u008 Ih8I 8I0 080d `IDD8 OIh0
N0d8 (vedinga). h0V 8I0 Ih0 80I0u00 O IIIu8 (kalpa) IO WDI0h
P0Om0IIV (sulba) I8 8II80h0d, DhOu0lI08 8ud DhOuOOPV (siki); 0IV-
mOOPV (nirukta); PI8Dm8I (vyikara1a); DIO8OdV (chanda); 8ud
88IIOuOmV]88\IOOPV (jyotii). NV II8u88ItOu8 O Ih0 u8D08 O Ih080
8C0u008 8I0 8DDIOXIm8I0 OI Ih0 DOuud0II08 O 80t0u00 8I0 uOI
ID0 88m0 80IO88 0IVIIZ8IIOu8. I 8hOud D0 uOI00 Ih8I OuI OuI O
Ih08080V0Dd08WI!h8uPu8P0. 1uIh080qu0,1 8h8uOIDOIh0I8DOuI
5 The verse: anantapiram kila sabdaitrall svalpalfl tathiyur bahava. ea vighniif
siral tato grihyam apiya phalgu halsair yathi k:fram ivibumadyit. The
proverb (which also appears as the fal metarule of a grammatical text:
Parihienduekhara 122); ardhamitralighavena putrotsaval manyate vaiyikara1i.
ARTIFICIAL LANGUAGES U!
lh0808uDdIVI8IOu8 DuI PIOuD Ih0m OO80VlOP0Ih0Iuud0I Ih0 8D0 O
IuPuI8II08 OI PI8mm8I.
1OW m8V W0 D0 8D0 lO 0mD0C 1udI8uIuPuI8lI08WIlhIu Ih0 0h8II
O 1IPuI0? 1I0m0IP0d 8P8IuIuuIu0I00uIh00uIuIV LuIOD08ud 8OIl
OOK888 I Ih0I00XI8I0d a tunnel, duPD0u08lHIh0 mOI0008IV VI8ID0
m8uI08I8IIOu8 O OIh0I 80I0u008 8ud Ih0I0OI0 uud0I00I0d DV Dt8IO-
II8u8 O 80I0u00. 800 IPuI0 . uI OOK8 8I0 Dt808dtuP. D8I W0
h8V0 h0I0 I8 8 0Ou8IIu0I du0 IO Ih0 u0P00I DOI OuV O IuCIP0DOu8
Mesopota
> 3000 BC
-.

m
/'
Gek
>600 BC
\
--
> 700 A -
I
Europea
> 1500 A
Aia
< - - M o d e r n -- --- >
Vec (< 5
t
c. BCE), P8 (5
t
c. BCE), Pajal (150 BCE), Koiki
(J c. CE), ... , Bha!oji it (1 J c. CE), NiojTbha!a (18
t
c. CE) ...
Wil, Chle (1808), A Grammar ofthe Sanskita Language, London
Bopp, Fr (1816), Oer ds Conjugationsstem de Sankitsprache, in
Vergleichung mit jenem de griechischen lateinischen, pesischen
und germanichen Sprachen, Fr
Teme, Pau (1982-83), ''e and For i te 'Gm a' ofPa,"
Studien zur lndologie und Iranistik 8/9:3-34

Figre 3. Lingistics: an apparent tuneL
U FRS STAL
08I0POII08 DuI 88O O IudIP0uOu8 8OuI008. h0I0 0XI8I8, Iu 80I, 8
0OuIIuuOu8 d0V0ODm0uI IOm Ih0 N0dI0 8V8I0m8 VI8 18uIuI I+!h
00uIuIV LL), Ih0 PI08I08I b8u8KIII PI8D8O8u, 8ud m8uV Olh0I8
Iu0udIuP 18I8Dj8I IJU LL) 8ud Ih0 Kisiki I!Ih 0. LL), uD IO
h8IIOjl 1II8 I!ID c. LL) 8ud 8POjlDh8II8 IIh 0. LL). h0
1udI8uII8dIIIOudIduOI8IOD!h0I0,DuIIII8Ih0uID8III W88DI0K0d uD
DV LD8I08 IKIu8 WhO80 b8u8KIII PI8mm8I OU I8 D880d uDOu
8nIuI8u DIIu0ID08. 1I tu8DII0d 1I8uZ ODD I) 8ud 0d IO Ih0
D0PIuuIuP O mOd0D IuPuI8II08. 18u DI0m0 IVZV) I08 Ih0
8IOIV Iu 8 0W D8P08 8I Ih0 D0PIuuIuP O 8u 8III00 Ou 18DIuI8
HI8mm8I. 1uUI8u PI8mD8I 0Ou!IIDuI0U uOI OuV lO mOd0Iu IuPuI8-
II08. 1I 8DI08d OV0I 8IP0 D8II8 O P8I8 Iu 8u 8DDI0d OIm 8IDI8I IO
Ih0 8DDI08lIOu8 O DDV8IC8 8ud OIh0I D88I0 80I0u008 Iu I00huOOPV.
1uCI8u IuPUI8II08 OIIPIu8I0d 8OuP I00II0I8 WhO W8ul0d IO DI0-
80IV0 Ih0II N0dIC h0III8H0 8Dd 8DDV II ID IIIu8. 1u0Ou00D0d WIID
m08uIuP, Ih0V 0Ou00uII0I0d Ou OIm 8uU Iu0OIDOI8I0d 8 POOd D08-
8uI0 O IuPuI8II0 8u8V8I8 Ih8I 0umIu8I0U IDIh0b8u8KItI PI8mm8I O
18uIuI IWhI0D IuCud08 m8uV N0dI0 OIm8 I00II0d IO 88 8u0h). O
uu00I8I8ud Ih0 European 8H0Im8Ih, !0I u8 I8K0 8 OOK 8I Ih0 8I08I
1udI8u PI8mm8II8u WhO 8COD8 OuI DII0 I8I. 8PO)lDh8!I8 W88
u0I!h0I 8 D8Ih-DI08KIuH 80I0uII8I lIK0 18DIuI, uOI 8 88VI8h OOW0I.
1I8 IuV08IIP8IIOu8 IuIO Ih0 18HIuI8u 8V8I0m W0I0 8ODhI8II08I0d 8I 8
OIDIddIuPVI00huI080V0.h0u1 WIOI0Iu Ih0 DI000dIuPD8I8PI0Dh
Ih8IIKIu8DI0K0UuD !h0II8dIIIOu, 1dIduOI ImDVIh8I h08I8II0d8I
Ih0 DOIuI Ih8I 8POjlDh8II8 h80 I080h0C. 10 W88 Iu uO DO8IIIOu IO
uud0I8I8ud 8POjl8 WOIK WhI0h LuIOD08D 8ud ^U0II08u 88V8uI8
D0P8u IO 888ImI8I0 OuV mu0h 8I0I. LIK0 U8uV OIh0I b8H8KIIII8l8,
IKIH8 08Iu0C lh0 D88I0 18DIuI8u I00huIqu08 Iu 1udI8 IOD 1udI8u
pandits OI 80hO8I8. m0 08D0d 0uOuR IO D0 8D0 IO WIII0 8 IVD0 O
PI8mD8I Ih8I LuIOD0 h8C uOI 800u D0OI0.
h0 8DDI08IIOu O DI0-18uIuI8u PI8mm8!I08 Iu8IPhI8 IOm 1udI8
Ih8I 8DI08d OV0I P8I8 I8 8 dIH0I0uI 8IOIV. 1I d080IV08 IDI00 80D8I8!0
800IIOu8. Ou0Ou Ih0Iu8IPhI8IB0m80V08, Ou0Ou!h0II8DI08CIO bOuIh,
bOuIh088I, L0uII8 8ud 88I P8I8, 8ud 8 IDIId Ou Ih0II 8IuI0 IO
08I8DI8h Ih0m80V08 Iu Ih0 08I L88I 8ud I080h LuIOD0.
!. The Vedic System of the Sounds of Language
Lu0 O ID0 PI08I08I dI80OV0II08 O lh0 N0dI0 I00II0I8 W88 ID0 8IlI0u-
8IOIV D880 OI Ih0 DIOdu0IIOu O Ih0 8Ouud8 O 8uPu8P0 Iu8O8I 88
Ih0V 8I0 m8uI08I Iu N0CI0 b8u8KIII.h0II dI80OV0IV DIOD8DV 8I8II0d
ARTIFCIAL LANGUAGES UV
with some of the regularities displayed by the square of twenty-five
consonants that is depicted in Figure 4. A reader without linguistic
training may understand what is at stake by refecting for a moment
on how to produce the sounds of the first row: k is produced by
pressing the back of the tongue against the roof of the mouth, c by
using the front of the tongue, t by using its tip to touch the teeth and
p by using the lips. In between come the sounds of retrofection,
produced by bending or fexing the tongue backward so that its tip
touches the spot on the palate that is marked with a capital 'T' with a
dot underneath. Thus, by going from left to right in the square of
sounds, we pass from the throat at the back of the mouth to the lips
in the front and that holds for each of the five rows.
If we go from top to bottom along any vertical colun, we move
from an unaspirate to an aspirate unvoiced consonant by adding air,
K c T T p

Kh Ch Th Th Ph

G J D D B

Gh Jh Dh Dh Bh

N\

N N N

vel pa!rl d labia
c
T
.
Figre 4. The Vedic system of the sounds of language.
U FRTS STA
repeating the sae for its voiced relative, and reaching at the bottom
the nasal of the same family which is created by passing air through
the nasal cavity while positioning the tongue and lips in. the same
places from left to right. It must have taken time to put all these facts
together and develop and extend the system by assigning to other
consonants, semivowels and vowels their place and articulatory
source. PaQini adopted 8 diferent order as we shall see.
Figure 4 depicts the sounds by letters, but the system arose in an
oral tradition. For ease of pronunciation, the consonants and semi
vowels were followed by a short -a. When scripts were developed later
on the basis of this system, they were generally syllabaries with a
stroke, curl or other special device added to turn ka into ki, ea into ci,
ta into ti, etc., another to tur ka into ku, ea into cu, ta into tu, etc.,
and so on for the others and similar
ly for the other consonants and
semivowels.
8. Indian Scripts of Asia
Once we know and understand the natural order of the sounds of one
language - a discpvery that seems to have been made only once - it is
not difcult to extend and adapt it to other languages that possess
diferent sound systems. This is exactly what happened in many parts
of Asia. We sometimes know about it from other sources but the
primary evidence is provided by the scripts themselves. Many Indian
scripts of Asia adopted the Indian system that developed from the
Vedic system of Figure 4. Modem students of these scripts who were
not familiar with the Indian system and only knew the haphazard
ABC's of the languages of the Middle East or Europe have often
ignored the system and paid attention exclusively to the shapes. A
typical example is the often reprinted handbook by Hans Jensen, Die
Schrit in Vergangenheit und Gegenwart. It denies that the Korean
script derives from the Indian because die Zeichenihnlichkeit ist
iiberaus gering ('the similarity between the signs is small') and must
therefore be the bewusste Erfndung eines einzelnen genialen Menschen
('the cons
c
ious invention of a unique genius': VV! 205). We shall
see, on the contrary, that the script was developed by a committee on
the basis of the Indian system and then expressed by a wholly original
system of shapes. The resulting combination makes the Korean
syllabary the most perfect that humans have evolved so far.
I refer to Asian scripts that adopted the Indian system as 'Indian
scripts of Asia'. Many of them adopted Indian shapes as well. Others
ARTIFCIAL LANGUAGES
1
accepted some but retained or modified others inherited from their
own traditions; or created new shapes, as did the Korean. Occasional
uses failed to make a lasting impact. Figure 5 is a rough map that
ilustrates the distribution of these varieties.
The Indian system together with its shapes was adopted by almost
all the scripts of South, Southeast and Central Asia. The South and
Southeast Asian scripts include Kharosthi (which incorporated a few
shapes from Aramaic), Brahm, Gupta, Khotanese, Npali, Nagari,
Bengali, Gujrati, Oriya, Pallava, Grantha, Tamil, Kannada, Telugu,
Malayalam, Sinhalese, Burmese, Thai, Ker, Javanese and
Balinese. The Central Asian scripts include again Kharosthi and
Khotanese in addition to Tibetan and lPhags-pa (which was created
from the Tibetan by the lama of that name for the Mongol Emperor
Khubilai Khan as an interational script for his Asian empire).
The organization of Chinese characters is so diferent from the
Indian that the latter had nothing to contribut. It caused confusion
Idian Systm
R Indian System
and Shapes
M Occaiona Us
of Indian System
Figure 5. Indian scripts of Asia.
1 12 FRTS STA
because Chinese Buddhists believed that each Indian shape was
independent and had its own meaning, like a Chinese character. But
there were a few exceptions. Hsieh Ling-yin (384-433), poet and
calligrapher, assisted by Hui-jui, a Buddhist monk, composed a
Sanskrit glossary in Chnese transcription in the Indian order
(Ziircher, 1959, Appendix to Ch. IV, note 1 25 (9) on p. 41 2). After the
ninth century, rhyme tables were composed for each tone which also
adapted that organization (van Gulik, 1980: 41).
As it happens, India also exported mantras. They conformed to the
Chinese belief because the moment one knows how to derive the
pronunciation of a mantra from the way it is written, one knows all
there is to know about it. That explains the success of the Siddham
alphabet, an adaptation of an Indian script, which became popular in
China and Japan though Sanskrit grammar itself never appealed to
the Chinese. Reacting diferently, Japanese and Tibetans were
familiar with the Indian system and arrived at a good understanding
of Sanskrit grammar ater centuries of intensive study. For Tibetan,
see Verhagen (1994, 2001).
In Japan, the Hiragana and Katakana syllabaries orignated during
the Heian period (794-1 185 CE). They clearly refect the Indian
system and gradually adapted it to the sounds of Japanese (see, e.g.,
Wenck, 1954-1959; Lewin, 1959). In the Heian period we find, for
instance, pa pi pu pe po, which later became fa.f fufe fo and during
the moder period ha hi fu he ho.
6
Hiragana is used for writing
indigenous Japanese words and some Chinese loan words; katakana
for non-Chinese (nowadays often English) loan words and technical
terms such as the names of animals and plants. The shapes are
simplifed forms derived from Chinese characters.
In Korea, Han-gul, the wo
r
ld's most perfect script, was developed
in 1444 CE by a committee of scholars appointed by the Emperor
Sejong. The underlying system is Indian, as we have already seen, but
the shapes of the characters are completely original and fully adapted
to the sounds of Korean. They refect the shapes of the mouth when
producing these various sounds (as does the European letter 'o'). The
committee report started with the basic insight, allegedly expressed by
the emperor: 'The sounds of our country's language are diferent
from those of China'.
6 I am grateful to Professors W.J. Boot and Michio Yano who have assisted me
and helped wt references to relevant literature.
ARTIFICIAL LANGUAGES
1 1 3
The study of the Indian scripts of Asia demonstrates that no
system should be followed slavishly. Careful adaptation is not only
more appropriate and useful. It conveys insight in the nature of the
sounds and other specific properties of a language.
,
9. The Arabic Alphabet
Controversy surrounds the Kitib al-'Ain or 'Book of the letter 'Ain',
composed in the 8th century LL by al-Kha1i b. Ahmad, teacher of
Sibawayhi, the most famous grammarian of Arabic.
7
Al-Khalil
re-arranged the letters of Arabic, starting in the back of the mouth
with the letter 'An followed by I, Ha, Kha, Ghain, Qaf, Kaf, et.
The standard alphabet returned to the original order, bringing
together letters of similar shapes that reflect an earlier stage in which
there were no diacritical dots so that, for example, the shapes of the,
2nd, 3rd and 4th letter, Ba, Ta and Tha, were the same (see Figure 6).
Al-Kha1Il was influenced by the grammatical schools of Kufan and
Basra of which fragmentary references remain since their teachings
were chiefy oral. These schools dealt with phonetic theory as well as
morphology and syntax. They discussed the classification of conso
nants and vowels, but seem to have had nothing to say about their
articulatory bases as far as I have been able to find out. Al-Khalil
may have come from Basra but he wrote his book with the assistance
of a native Iranian called Laith in Korasan, the eastermost part of
Iran which is also the gateway to India (Haywood, 1960: 37).
Recent authors who deny that there was Indian infuence on the
Kitib include several scholars writing in Arabic referred to by
Danecki (1985) whom I have not been able to consult; and in addition
Law (1990) and Versteegh (1993). Those who accept Indian infuence
include Haywood (1960), Wild (1962), Danecki (1985) and Carter
(1990). Those who deny that there was any infuence argue, frst, that
there is no evidence, meaning texts, for transmission of that
knowledge from India to any Arab country; second, that there is no
evidence for contacts between Arab and Indian scholars at the time of
al-Kali and Sibawayhi; and third, that there are many diferences
between Indian theories and those of Al-Khali and Sibawayhi who
also difered from each other in their understanding of the bases of
articulation.
7 I a grateful to Professors Oskar von Hiniiber, Richard C. Martin and Kees
Versteegh who helped me with references to relevant literature.
1 14 FRTS STAL
Arabic: standard alphabet i which letters of similar
shape ae brougt togeter:
;
b t t
j
h k d . r z s s s
.
y
u 6
( t
;

.) .
,
I
(
d t z
c
g
f
q
k I m n w h
y

J J j
t
t
J
3
. J ) 0
)
;
I
Sequence intoduced i the Kitib al- 5yn by al-Kalll b.
Aad (78/19-91 CE), teacher ofSibawayhi, autor of
te frst flly develope gam a of Arabic. Allegedly
witten i Koraa:
c
h h k
g q
k
j
s d s s t z

t (
;
t
t
3
.

.
J !
_
J
(
d t z
. t r I n f b m w
,
y

;
u
j
:
6 .
J J
J
y
)
)
I
Figure 6. The Standard Arabic Aphabet and the Indian Aphabet of the Kitib
al-'Ain.
What weakens the force of the fst argument is that it is widely
agreed that the hypothesis of Greek influence on Arab grammar,
whether directly or through Syriac intermediaries, apart from the fact
that the Greeks did not have much to ofer in this field, also lacks
documentary evidence (e.g., Carter, 1 990a: 109, 1 19). As for the
second, around the time al-Khall was writing, before the 'Abbasid
period and at the beginning of Arabic science, Indian astronomy and
mathematics played an important role. In
7
35 CE, a Zj al-Arkand
based on Brahmagupta's Ka14akhidyaka was composed in Sind; in
ARTIFICIAL LAGUAGES 1 1 5
!+Z, 8 Zfj al-Harqan was 0OmDO80dDuId0IIV0d IOm

IV8Dh8I8, 8ud
Iu lh0 08IV 770's, Ih0 Zfj al-Sindhind al-kabfr W88 D880d Ou 8uOIh0I
WOIK O I8Dm8PuDI8 |1IuPI00, 1976, 1 51-69). h0 1udI8H II8dI-
IIOu 8DI08d quI0KV IhIOuPhOuI 188DI0 8ud8 OIuPI00 1 968, 1970)
8ud 8-bh8 , WhO W88 8 PI08I 80I0uII8I WIIh 8 K00u 8u8VII08 m
Iud,
WOud 00II8IuV h8V0 D00u 8mII8I WIIh II8 0XI8I0u00. P8 OI 1I8u,
Wh0I0 1udI8u KuOW0dP0 8ud 80I0u008 D8d OuP D00u 8DII8I 8ud
Wh0I0 1udI8u DOOK8 OuPI8DD8IOI vyikarala, KuOWu88 fy'kr, h8d
D00u K0DI 8OuP WI!h OIh0I 0888I08 Iu Ih0 IOV8 II008uIV 8Iu00 Ih0
b8888uI8u D0IIO0 |d0 N0u8800, 1949: 2), Wh8I 0Oud 8-1h8D uOI
h8V0 DI0K0d HD ID0I0?
h0ID1Id8IPum0uI I8Ih0DO8I IDI0I08IIuP8u0d00ud0d IuPI08I08I
U0I8I DV NIV0u L8W (1990) WhO h88 I8K0u Ih0 IIOuD0 IO 8IudV Ih0
1udI8u Pritisikhya II0I8IuI0 WDI0h I8 DO8IV 8V8I8D0, 88 h0 IIPhIV
uOI08 (223, note 2), Iu uPI8h II8u88IIOu. 10 h88 0OH0lud0d Ih8I
Ih0I0 8I0 8 PI08I m8uV dI0I0u008 Od0I8I D0IW00u 1uUI8u 8ud PI8D
I0XI8. 1I0I0uC08 8I0, 08 W0 h8V0 ju8I 800u, DI00I80V Wh8I Ou0 WOud
0XD00I. I8DI0I8V0IV dI0I0uI IOmb8u8KIII. 1I I88I-0I0D0d 8uVW8V
IO 888um0 Ih8I P-1h81 8!udI0d b8u8KIII II08II808, DuI h0 W08 V0IV
V0I88II0, WIO0 WOIK8 Ou8VuI8X, DIO8OUV 8ud mu8I0, 8ud II I8 h8Id IO
D0I0V0 Ih8I h0 dId uOI uud0I8!8uU Ih0 1udI8u 8V8I0m Wh0u h0 D08I0
8DOuI II, 88 dId m8uV OID0I8 D0OI0 8u0 8H0I DIm. 10 DuI Ih0 0II0I
`PIu, WDI0h dO08 uOI O00uI Iu b8u8KIII, 8I !h0 D0PIuuIuP D008u80 II8
OIIPIu I08 d00D08I Iu Ih0 lhIO8I. 10 W88 uOI 8V0I80 IO u8IuP
0XD0IID0uI8 m0IhOd8 IOI Ih0I0 I8 8 8IOIV 8DOuI hIm DulIIuP hI8
HuP0I8 Iu hI8 mOuID IO HDd OuI hOW 8Ouud8 8I0 DIOdu00d. 1II8 IK0V
!h8I 1udI0u8 dI0Ih0 88m0 DuI Ih0I0 I8 uO I0XIu8 0VId0u00 lO DIOV0 II.
L8W8 8IPuD0uI I8 D080d uDOu 8 80II08 O 0O80V I08I0d
888uDDIIOD8!
It is virtually without preedent for a system- an abstraction, afer all - to be taken
over without any of the associated doctrine. In those cases where the borrowing of a
linguistic system is historically tested - Romans from Greks, Jews from Arabs,
European veracular grammarians from Romans, Japanese and Chinese from the
West (to name only a few) - it has invariably been acompanied by the wholesale
borrowing of concepts, terms (loanwords and calques) and even examples, as well as
by frequent attempts to fnd a new function for redundant doctrine or to accomodate
recalcitrant native phenomena within the foreign system. In short, the borrowing of
isolated details along with a foreign system is historically well-attested, whereas the
borrowing of a system without any of the details on which it rests is almost unknown
(Law, 1990: 216).
1uDII0, Wh8I L8W d080IID0888`D18IOII08VW0-8II08I0dI80X80IV
Wh8I 1 080d `88VI8h IDII8IIOu. h0Ih0I OI uOI II 8DDI08 IO Ih0
1 16 FRTS STA
examples he cites, we have seen that in most of the Indian scripts of
Asia, the system was not followed slavishly with all its details but
adapted to the language that adopted it because the languages were
dif erent. All these other adaptations strengthen the case for adap
tation by al-Khall. As for the diferences, the nail was hit on the head
almost DHa century ago by Haywood (1960: 37): 'A comparison of
his order with that of the Sanskrit alphabet shows sufcient broad
similarity to suggest infuence, yet enough divergence in detail to
indicate an independent mind moulding borrowed ideas'.
That there is no written evidence for such borrowing apart from
al-Khalil's alphabet itself simply shows that the channels were oral. It
is not diferent from many other Asian borrowings of Indian scripts
or from the case of Ibn Sina (Avicenna) of whom it is said that he
learned Greek philosophy and Euclidian geometry by himself (that
is, from manuscripts or books), but picked up Indian numerals from
his grocer (Nogales, 1990: 399). I conclude that in gramr as in
mathematics, ideas that are part of an oral tradition may be picked
up by others who do not slavishly follow texts but understand the
subject.
10. Paradoxes Surrounding the Transmission of Linguistics
and Mathematics
From the Indian point of view, the history of science does not look
like the Needhamian rivers of Figure 1 or the tunnel view of Figure
at all. It rather looks like a tree with its roots in India and its branches
spreading out over the Asian continent as depicted by the map of
Figure 5. The Indian system spread orally, or to be more precise;
through the ears, minds and mouths of merchants, monks and other
travelers. Oral knowledge spreads more easily than written knowl
edge especially when it is knowledge of sounds. This brings us face to
face with several paradoxes. The first is that the Indian system did not
originate in spite of the absence of writing but because of it (Staal,
1989). The second is that its greatest applications were not coned
to oral traditions, but pertained to the scripts of Asia, i.e., writing.
The numerical expressions and other scientific traditions that Michio
Y ano studies elsewhere in this issue (141-1 58) are partly oral and
partly written. Given the historical and conceptual priority of the
study of language in India, Indians were predisposed to thn of
numerical expressions as language. It may explain another paradox,
viz. , that Indian mathematicians preferred orally based systems such
ARTIFICIAL LANGUAGES 1 17
as the ka(apayidi in stead of making use of the numerals for which
they were famous. Fortunately for the moder world, numerical
expressions were adopted by the Arabs. Unfortunately, the Indian
sound system did not take root in the Arab world and t
h
e moder
world did not develop rational alphabets and linguistics as early and
well as it did mathematics.
I mention these facts and paradoxes although they deserve much
more discussion and may stand in need of qualification. They may
help explain that the earliest artifcial language was oral and did not
arise in the context of mathematics but in linguistics.
I1I EARLY ARTIFICIAL LANGUAGES
1 1 . Piini's Grammar of Sanskrit
PaJini (4th century BCE) must have been familiar with what I have
referred to as the Vedic sound system, depicted by Figure 4, but he
ordered the sounds of Sanskrit diferently and, from our point
of view, counter-intuitively. To gain a general idea of the significance
of such counterintuitive expressions we should recall that historian of
science Aexandre Koyre taught that the world is based upon
principles that we cannot intuitively grasp. We shall retur to some of
the reasons for this remarkable fact, but in the present context it is
relevant to note that Koyre was thinking primarily of the physical
universe and not of the study oflanguage. PaJini' grammar, however,
fully supports such a view. It shows that language is as mysterious as
the rest of the world and illustrates that being human does not imply
that everyhing that is human is immediately given and obvious to us.
We shall see that Pa:ini needed an artifcial language in order to
express te counter-intuitive features of natural language. To
understand his artificial expressions themselves we should take a
closer look at the problems he faced.
PaJini used rules (siUra) to express the facts of sandhi or 'euphonic
combination'. The genre of sutra is the preferred form of Sanskrit
expression used in the ea
r
ly Indian sciences (enumerated at the outset
of Section 6, Linguistic: An Apparent Tunnel): a sutra is a concise
formula-like expression (Renou, 1963; Staal, 1992). It is in prose,
unlike, e.g., the poetic expressions used by some later Indian math
ematicians (the Vedic geometers, who were contemporaries ofPaJini,
used sutra prose). The subject is familiar to readers of the Joural of
Indian Philosophy, who may also refer to an earlier and more
1 1 8 FRTS STAL
Sanskritic article in this Journal (Staal, 1995a) where the demon
stration that Patini created an artificial language is given in greater
detail than in the slightly idealized form in which it will bt presented
here.
What sitra rules are like will become apparent soon, but I must
fst explain what is sandhi. Fortunately, sandhi does not occur only in
Sanskrit and so we may use illustrations from English and French.
Both languages are, like Sanskrit, replete with sandhi, but the large
majority of cases is not refected by spelling and, therefore, easier to
explain orally than in a written paper.
An example from English that is not reflected by the spelling
concers the vowel of the definite article the which is pronounced
diferently - at least in some English dialects - in its two occurrences
in 'the coconut' and 'the apple'. The fst -e- sound sounds like the -u
in but and is due to the fact that a consonant follows; the second
sounds more like -ea- in please because a vowel follows. Both facts
must be described by sandhi rules.
The example from French is reflected by the French spelling. If il
a 'there is' is turned into a question, it does not become *a il? but
a-t-il? 'is there?' The insertion of -t- must be described by a rule of
sandhi.
Sanskrit spelling always refects the sounds of pronunciation. The
reason is that the analysis of language was based upon the spoken
language and was itself oral, not only among the pre-Pa:inian
grammarians but probably including Patini himself (see Staal, 1986
with which von Hiniber, 1990 and Falk, 1993 are consistent). That
had changed by the time of the commentator Patafijali (c. 1 50 BCE),
but by then there existed already a sound foundation for linguistics
and for a scientific organization of scripts as we have seen in
Sections 8 'Indian Scripts of Asia' and 9 'The Arabic Alphabet'.
My illustration of a sandhi rule in Patini's Sanskrit gramar
describes the process of retrofexion. For retroflexes the reader is
again referred to Figure 4 where they are marked with a dot under
neath and listed within the rectangular box of the central column.
Sanskrit possesses two other retroflexes whch are not in the square
but articulated in the same manner and one of which is also marked
with a dot in the transliteration. The first is the semivowel r, which is
more like the French or Italian than the English r. The second is a
fricative or sibilant, written or, in popuar transliteration 'sh' as in
'Vishnu'. The reader will be able to produce it like any other retroflex
sound, by bending or flexing the tongue backward so that its tip
ARTIFICIAL LANGUAGES 1 19
touches the spot on the palate that is also marked with a capital
'T' with a dot undereath in Figure 4.
The process of retrofexion is complex. Modem gram ars of
Sanskrit devote much space to it and it is a notorious stumbling block
to beginning students of Sanskrit. Pa:ini wanted to explain the rea
sons and express them with the help of a short and intelligible rule
based upon general principles. He was able to do that though he
needed to add a number of special cases and exceptions. I shall explain
his rule, but simplify the complexities of the exceptions by grouping
some of them together, following Pa:inian principles though not the
letter of the master. It will enable us to see in the next section and in a
relatively simple manner how Pa:ini achieved his goals by introducing
artificial expressions as well as an artificial language.
Pa:ini's rule (8.4. 1) is simple:
Mter r and , n becomes 1 in the same word.
(
1)
'In the same word' i s added because this kind of retrofexion does
not cross word boundaries. I shall omit it because it is not part of
the artificial language. Many of the exceptions are captured when
we add to (1) the following restriction which I shall later formulate
as an artifcial expression in Pa:inian fashion. It occurs elsewhere
in the grammar but Pa:ini does not use it here for reasons of his
own:
unless c, (, t, eh, th or th interfere. (2
)
Here are three examples:
A. No consonant occurs between r and n and since nothing inter
feres; retrofexion takes place: kar-na becomes kar!a 'ear' .
B. Consonants interfere but not the ones mentioned in (2). The famous
Indian epic * Rimiyana refers to the 'comings-and-goings' (iyana)
of Prince Rama. The 'r' that causes retroflexion stands at the
beginning. Several syllables intervene but do not interfere and the
fnal 'n', therefore, becomes retrofex: Rimiya!a.
C. c, one of the consonants mentioned in (2), interferes in arcanam,
'homage', and the retroflexion caused by the 'r' is blocked. We
thus arrive at the correct form which is: arcanam, not arca!am.
12. Pi!ini's Articial Language
Let us begin with a closer look at rule (2). In the table of Figure 4, it is
not easy to single out the consonants that have to be enumerated-c, (,
120 FRTS STA
t, eh, h and th-by making use of general rules or principles. They occur
in a rectangle that could be cut out from the table, but the table was
orally conceived, understood and transmitted; not fit for cutting and
pasting. Simlarly, consonants that do not interfere, such as k and p in
the fst row, do not occur together; they occur at opposite ends. Being
orally conceived, the table possessed a feature that facilitated pro
nunciation, recitation and recollection as we have seen: the consonants
are not mentioned by themselves as k, e, t, etc., but pronounced with a
following short -a, i.e., as ka, ea, a, etc. Palini was aware of the fact
that ka is the name of the consonant k and belongs to the metalanguage
of grammar. The concept was familiar in the linguistic tradition from
Palini, who distinguished between use and mention (Brough, 1951), and
Pataijali, who used the technical term paribhii or 'metarule' when
referring to rules about rules (Staal, 1975).
Various combinations of sounds are required to formulate the
numerous sandhi- and other rules that the grammar needs. Pilini,
perhaps with the help of assistants and students, studied and collected
all these combinations and arrived at a diferent system for ordering
the sounds of Sanskrit. It is not two-dimensional, like the square of
Figure 4, but linear like natural language itself. The system includes
metalinguistic markers that pick out the combinations he needed.
This resulted in what were later called the
S
iva-sutras:
a i u " I r l K I e o N I ai au C I ha ya va ra T I la " I ia ma Jia !a na
M I jha bha N I gha rha dha I ja ba ga ra da S I kha pha eha ha
tha ea a ta V I ka pa y I sa a sa R I ha L I I (3
)
In this written representation ofthe sequence, the sounds of Sanskrit
object-language have been expressed by small letters. They start with
vowels and continue with semivowels and consonants, each followed
by the short a. I have used capitals to indicate the metalinguistic
markers of the metalanguage. Their use is explained by a metarule:
An initial sound joined to a fnal (indicatory) sound
(denotes the intervening sounds as well).
(
4
)
Thus, aN refers to a, i and u, aC refers to any vowel or diphthong,
haL to all consonants (though ha occurs twice), etc. What we need is
ehaV which refers to precisely the exceptions listed in (2). It refers to
consonants articulated in the central area of the mouth (see Figure 4)
that interfere with retrofexion, unlike others that do not - e.g., k and
p that are produced at the periphery, in the throat and by the lips.
ARTIFICIAL LANGUAGES 121
Kiparsky (1991) has shown that the main principles of Pa:ini's
grammar can be inferred from the listing of (3) and the artificial
expressions that are formed from it with the help of ( 4).
Do expressions like a, aC, haL and chaV constitute what I have
called 'the artificial language' of Pa:ini's grammar? They do not
because an artificial language is more than a set of artificial expres
sions. It is a language that combines these expressions together into
larger expressions like sentences or statements in accordance with
syntactical rules. In that respect, artificial languages are like natural
languages: they posses a syntax. They also posses, like natural lan
guage, a semantics. What is the domain of semantics of Pa:ini's
grammar? The answer is easily given: it consists of the forms of the
Sanskrit object-language. What is its syntax? That is our next
question.
Let us begin by taking a look at the general structure of a gram
matical rule such as (1). Many of these rules say that A becomes B or,
equivalently, that B has to be substituted for A (both interretations,
the 'diachronic' and 'synchronic', were distinguished from each other
before Pa:ini). Modern systems of linguistics sometimes express that
relationship by an arrow as in: A - B.
Pre-Pa:inian grammarians had already adopted a convention: to
refer to the A (in any of these expressions), they attached to it the ending
of the Nominative case in Sanskrit; and for B they attached to it the
ending of the Accusative case. In other words, they used the case
endings of the object-language metalinguistically. They expressed that
convention by a metarule.
Pa:ini introduced four metarules. The first is semi-artificial: it
states that the subject of the rule, i.e., the element that is substituted,
is expressed by the Nominative Case. The metalinguistic uses of three
other Cases are laid down by three artificial metarules that refer to
the Genitive, Locative and Ablative Cases:
The Genitive case ending is used for that in the place of
which (something is substituted); (5)
when something is referred to by the Locative ending
(the substitute appears) in the place of preceding element;
(
6
)
when (something is referred to) by the Ablative ending
(the substitute appears in the place) of the following.
(
7
)
Applying these rules, Pa:ini formulates (1) as
122 FRTS STA
r and + Ablative ending, n + Genitive ending,
v + Nominative ending (8)
which may be read as
After r and ' in the place of n, substitute v .
(
9
)
There is no scope for applying metarule (5) but (8) is the general
format of statements in the artificial language that occurs in many
hundreds of the approximately 4000 rules of the grammar.
Palini's artifcial language, therefore, is not the language of his
entire grammar. That is not diferent from other uses of artifcial
languages. Aristotle's artifcial language consists of statements such
as A, B, C and D that are logically related to each other; but they are
generally embedded in Greek phrases translated as 'Therefore it is
clear that A', 'But we may prove that it is true that B', 'Thus we have
established the distinction between C and D'. Einstein' s famous
article of 1 91 6 which introduced general relativity is riddled with
formulas and equations but their relationships are similarly expressed
through natural language as in: 'Because of the symmetry of A with
respect to the indices B and C the third expression on the right
vanishes provided D is an asymmetrical tensor as we shall assume . . . '
In both cases, A, B, C and D are statements of an artificial language
and the rest is Greek or German. Complete artificiality and formal
ization was sought in nineteenth century mathematics and, in logic, in
Frege's Begrif schrit.
One characteristic of Pa:ini's artificial language is the linguistic
zero, or rather: zeroes. It may be illustrated by constructing a frag
ment of a Palinian grammar for English. In English, the formation of
the plural may be described, to begin with, by something like (10),
where P is the plural marker:
Noun + P -- + noun + sufx (e)s ( 10)
Additional rules or specifcations have to be added to express that -s
follows vowels and -es consonants. But there are exceptions which
Pa:ini would treat with the help of zero sufes, e.g.:
If the noun is sheep - s = 01 (zero suf which causes no change)
if the noun is man - s = 02 (zero sufx which causes man to change
into men)
if the noun is women - s = 03 (zero suf x which causes woman to
change into women)
ARTIFICIAL LANGUAGES
123
if the noun, is mouse s = 04 (zero sufx which causes mouse to
change into mice)
To sum up: PaJini' s grammar makes use of articial expressions
such as a, aC haL and chaV and an articial language that is
constructed from these expressions in accordance with metarules
adopted from Sanskrt syntax.
13. Notes on Aristotle's Articial Language and on Variables
The basic expression of Aristotle's logic is: 'b belongs to (ncQXEt) a'
/ G
or 'a is b', e.g. , 'the horse is white'. Here 'horse' is subject and 'white'
is predicate; it is therefore a predicate logic (unlike the logic of the
later Stoa which is about propositions). Aristotle's logical language is
not just a set of artificial expressions such as a or b, but a language
consisting of statements or rules - in other words: an artificial
language.
Aristotle' s artificial language introduced quantification into these
predicate expressions by constructing four basic forms:
'b belongs to all a' or 'all a is b' - later: Aab
'b belongs to no a' or 'no a is b' - later: Ea
'b belongs to some a' or 'some a is b' - later: lab
'b does not belong to some a' or 'some a is not b' - later: Oa
The system was fully developed and further enriched by intro
ducing modalities (ossible, necessary, . . . ). There exists a large lit
erature about it, including attempts at formalization.
Aristotle was a contemporary of PaJini and added something
that PaJni did not have. The expressions to which I have referred
as a and b are variables, one of Aristotle's greatest inventions
(Lukasiewicz, 1957: 7). Variables are similar to pronouns in a nat
ural language like English: 'If anyone wants food, he should go to
the kitchen; if he wants drinks, he should ask Johnny' . By 300
BCE, the Latr Mohists in China also knew variables which neither
Plato, nor early Sanskrit or Chinese or the later Chinese Buddhist
logicians used (Harbsmeier, 1998: 333, 406). All these others may
use ' such-and-such' for an unnamed entity, but if they use it twice,
it may refer to diferent entities. Variables occur also in Chinese
legal texts of roughly the same period as the Mohist logicians, e.g. :
'X(chia) and Y(i) do not originally know each other. X goes on to
rob Z (ing). As he arrives, Y also goes along to rob Z and speaks
with X'.
1 24 FRITS STAAL
I conclude that artificial languages appeared in human history in
India, Greece and China around the same time. It is sometimes
referred to as the Achsenzeit, 'the axial age', a period without exact
boundaries or demarcations. It is not liable to a simplistic explana
tion, e.g., 'nomads settling down to agriculture' - a practice known to
ants who do not only have queens and slaves but the slaves are
engaged in cutting and collecting leaves. How does the appearance of
artificial language look within the wider perspective of the develop
ment of human language? We shall try to answer that question in
Section 1 5 'From Natural to Artifcial Language' afer sketching an
illustration in Section 14 'From Predicates to Functions' first.
I FROM NATURAL TO ARTIFICIAL LANGUAGE
14. From Predicates to Functions
We have seen how Paiini used the case-endings of Sanskrit
metalinguistically, making their endings the most important
syntactical device in his artifcial language. Sanskrit is rich in cases
and case-endings. Cases and case-endings occur in some languages
only, but case-like structures are widespread. Aristotle constructed
his artifcial language by starting with a case-like structure, viz., the
distinction between subject and predicate. Its expression is clear in
Greek and the subject/predicate distinction seems to be a language
universal, that is, it exists in all human languages. The subject
predicate structure, however, does not go very far. Wat can a
subject-predicate structure not do?
A notorious example comes from one of the greatest builders of
artificial expressions and languages. Leibniz tried to analyze 'Titus is
bigger than Caius' as a conjunction of two subject-predicate state
ments: 'Titus is big in as much as Caius is small'. Historians of logic
Kneale and Kneale comment: 'The strenuousness of his eforts to
preserve the old theory that every statement ascribes an attribute to a
subject shows his own uneasiness, and has stimulated later logicians
to shake themselves free from this part of the tradition' (Kneale and
Kneale, 1962: 324-325).
'The old theory' to which Kneale and Kneale refer is Aristotle's
theory, not only the logical portion that refers to subject and predi
cate, but including the epistemological and metaphysical part that
refers to subjects or substances and attributes or qualities. Both
ARTIFCIAL LANGUAGES
125
Aristotle and his Indian counterparts in the Nyaya-Vaiseika tradition
started their logic with subject-predicate expressions and transformed
that system into metaphysics by imagining that the world could be
described by statements that ascribe attributes or qualities to subjects
or substances. The Greek and Indian stock examples are, respectively,
to ascribe white to horse and weight to water. A brief look at a
textbook of physics, astronomy, economics, genetics, linguistics or
any other science shows that it does not work.
What should Leibniz have done? A simple solution would be to
replace the subject-predicate relation, which makes an assertion
about one term, by another that makes an assertion about two; in
other terms, replace a monadic by a dyadic relation. Let me try to
make it a little more precise and illustrate how a powerful and flexible
artificial expression may have developed from the subject-predicate
structure of natural language. If not 'developed from', it might at
least have been suggested by or is simply related to.
Let us start with 'the horse is white' or 'water is heavy'. Here a
predicate b (white or heavy) is predicated of a subject a (horse or
water). Let us write it in more formal terms as b(a). This expression is
inspired by logic where it is often written as fx) read as 'x isf'. Why
not a(b) or x() or something else? Because, in these simple monadic
cases, the entire expression or proposition predicates something of
the subject. For example, negation of the proposition is the same as
negation of the predicate and not of the subject. Arstotle explained
it: 'The negation of man is is man is not, not not-man is' .
Following Whitehead and Russell's Princiia Mathematica, mod
ern logic uses a symbol for negation such as '-'. A proposition of the
form fx) is negated by placing that symbol imediately in front of it,
where it also occurs immediately in front of the predicate:
r (x))
=
( r J(x)) . ( 1 1)
This identity entitles us to omit parentheses and write, without
ambiguity '-fx)' . Thus we have a good reason, for the particular
form of '(x)' itself- an illustration of the close relationship between
Aristotle, modern logic and natural language.
The expression 'jx)' has the same form as a mathematical
function:
y
=
f
x) . ( 12)
In mathematics, the function 'f' expresses a relation between two
objects which specifies, for each x, at most one y, the value of fix). In
126 FRTS STAL
the language of functions, it is easy to express what subject-predicate
structures cannot, e.g. a dyadic structure:
J(x, y) ( 1 3)
e.g., 'Brutus killed Caesar'; or a tryadic structure:
f(x, y, z) ( 14)
e.g., 'the fy moves from A to B', or 'John gave the book to Mary'; or,
generalized:
J
(xl
,

, x
n
)
( 1 5)
with (13
)
and (14) as special cases for n = 2 and n = 3.
Natural languages do not introduce infinity into the subject
predicate framework but use recursive devices such as 'and' to enu
merate subjects, predicates or other expressions, e.g. :
Gopal is brilliant, far-sighted, and mean, ( 1 6)
which corresponds to
J(x) & g(x) & h(x) ,
( 1 7)
or
Amir, Ka
r
in and Yuki are forgetful, ( 1 8)
which corresponds to
f(x) &J(y) &J(z) . ( 19)
The mathematicians who introduced expressions for functions
were speakers and writers of one or more natural languages and
familiar, at least implicitly, that is subconsciously, with some of their
structures. In explicit terms, they were thinking of numbers or points
as variables for x, y, etc. The history of some of these expressions is
relatively well known. Much of it took place in the eighteenth cen
tury. Expressions such as (12) are due to Euler, 'the most successful
notation builder of all times' (Boyer and Merzbach, 1991 : 442). They
had been used earlier without parentheses by Johannes Bemouilli I
for the functor <. The letters f, < and \J came in common use through
the work of J .L. Lagrange. Leibniz was the frst to use subscripts for
functors.
The mathematical concept of function is much ri
c
her and deeper
that these simple cases illustrate. Functional analysis covers a good
part of mathematics and an even larger part of the vast body of
mathematics that is used in physics. The successful use of such formal
mathematical expressions suggests that they tell us something about
ARTIFICIAL LANGUAGES 1 27
the universe. Even so, these notations and expressions are related to
forms of natural language and no animal without language is likely to
have been able to discover or invent them.
15. From Natural to Articial Language
Our remote ancestors developed natural language some 70-
50,000 years ago; our more immediate predecessors developed arti
fal language a few thousand years ago. It explains that we are
relatively well informed about the origins of artificial language in
diferent civilizations and regions on the planet, but know little about
the remote origins of language itself The lack of balance in our
information reflects our present perspective, the only perspective we
possess at the stage of development we happen to have reahed at the
present point in time. It is a diference that may blind us to the fact
that the two types of language are just two steps in the same
direction. But are they?
In order to find out we should look at the diferences and simi
larities between the two varieties. One diference seems to be that
artificial languages were deliberately designed and mostly for a def
nite purpose: the expression of knowledge. (PaQini's grammar and
Aristotle's logical works were systematic treatises, not teaching
manuals; they became normative with the passage of time.) Natural
language, on the other hand, seems to have been selected and grown
in response to the need for communication, the exchange of infor
mation between senders and receivers. The main evidence for that
view is that we share senders such as mouths and receivers such as
ears with many other animals who use them for the purpose of
communication. Human language is constrained by the structure of
these organs and by a computational system in the human brain that
may be diferent in important respects from the systems used in
animal brains, which themselves vary enormously from species to
species.
8
A diference that is only apparent is that both kinds of language
look diferent. Expressing it thus presupposes that both varieties are
8 This is a large and controversial topic. Noam Chomsky is the main proponent of the
idea that language is primarily for knowledge, not communication. His is a statement
about function, not evolution. See, e.g., Chomsky (2002), Hauser et al. (2002) and cf.
Staal (200lb). Ancient Chinese thinkers of various persuasions defined language i
terms of communication (Harbsmeier, 1998: 47-48). Indian philosophers oflanguage
such as Bhart,an use the term vyavahira, practice, usage, communication to char
acterize the use of ordinary language (see, e.g., Houben, 1995: 333, note 522).
128 FRTS STAL
written. Natural languages are often written, especially the important
ones; but even in the present world, the majority of languages are not.
None were written before the invention of writing, a few millennia
ago. The same applies to artificial language: it is now invariably
written, often in computer versions where 'written' itself is a meta
phor; but Palini's grammar was not written and the earlier Vedic
treatises on language were neither. The phonology of Palini's oral
artifcial language is the same as Sanskrit phonology. The
corresponding property of the artificial languages of logic and
mathematics is well-formedness.
The most signifcant similarity between natural and artificial lan
guages is that both varieties consist of syntax and semantics. The
most important properties of syntax are that it is able to make infinite
use of fnite means and that there may be more than one semantic
interpretation of a syntactic expression. In natural language, the
latter property is often called ambiguity, a notion that is surrounded
by an aura of poetic metaphor and creativity. Artificial languages are
also creative but in mathematical physics, something even more
mysterious may happen: a syntactic expression may agree with
empirical data though diferent interpretations are assigned to it, and
sometimes no interpretation at all. A famous example of the latter is
Planck's formula E = hv which was in accordance with certain
measurements, but seemed to make sense only much later after
quantum mechanics was born (Dyson, 1 988: 21 ; Farmelo, 2002a, b).
A earlier example is imaginary and complex numbers: they do not
make sense - for how can the root of -1 exist? But the unverse
appears to love them.
That we should be able to understand the universe is, of course,
not obvious. And why should we be able to understand it in terms of
our human language that was originally designed for practical com
munication amongst ourselves? Is artificial language really a better
instrument since it was created for the express purpose of knowledge
and understanding? We shall retur to this question but conclude for
the time being, that there are diferences as well as similarities
between natural and artificial languages. The paradox about artifcial
languages is that they are, in their original intent and ideal form,
independent of natural language; but their origns and historical
development have been inspired by it as we have seen. That latter fact
is not surprising because it is unlikely that they fell from the sky. Is
the artificial variety, then, a mere extension of (art of the natural?
Or is it something basically diferent?
ARTIFCIAL LANGUAGES 129
16. Conclusions and Questions
First conclusion: the spread of articial languages
Our frst conclusion was that artificia1 languages exist across sciences
and civilizations as my title
c
laimed. We found that their earliest
occurrences were not connected with mathematics as are their more
recent and well known forms; but with linguistics and logc. Artificial
languages have continued to fourish and have been adopted in many
other sciences and disciplines.
Second conclusion: linguistics in India
Our second conclusion was that the revolutionary successes of lin
guistics in India were largely due to the use of an artifcial language.
Does the same hold for other sciences?
Third conclusion: physics and mathematics in Europe
At the end of Section 4 'Rivers Flowing into the Sea' we touched
upon this question in relation to the European scientifc revolution
in mathematics and physics. We saw that Newton's Latin was often
unclear. Actually, the formulas that are referred to as 'Newton's
equations' were introduced later by Eu1er, Daniel Bernouilli and
other mathematicians. It is true that we, today, can easily read
them into Newton's words, but we do so by hindsight (Truesdell,
1968: 167). I concluded that the so-called 'European' revolution
was not a product of Europe or of world history, but was triggered
by the introduction of formulas and equations that were easy to
understand, independent of the varieties and vagaries of natural
languages, and led to a language that is universal (Staal, 1995a,
2003b, 2004). My thrd conclusion is that the revolution in physics
and mathematics that took place in Europe was a revolution in
language.
Is that what Galileo had in mind when he wrote that the uni
verse is a book written in the language of mathematics (full
translation in section 3 'The Principle of Modernity')? He did not.
The idea that the universe is a book is a common topos throughout
the Latin middle ages (Curtius, 1948), a refection of the mono
theistic religions of Judaism, Chistianity and Islam, the 'reli
g
ions
of the book' as the Qur'an calls them. Galileo referred to such a
book, but its characters are not letters: they are 'triangles, squares,
circles, spheres, cones, pyramids, and other mathematical fgures'
(Blay, 1998: 1). He was not thinking of algebra, that 'barbarous' art
130 FRTS STAL
as Descartes called i
e
- ironically for the founder of analytical
geometry which established a link between geometry and algebra.
Like Newton, Galileo continued to work in the Euclidian tradition.
His metaphors support the idea that the European revol
u
tion was a
geometrical revolution, not inspired by the language of algebra and
certainly not a revolution in language. Galileo, Newton and even
Descartes failed to understand the power and impact of the alge
braic language that was to be the language of the future.
Fourth conclusion: linguistics, mathematics and physics
My fourth conclusion is a generalization of the second and third:
revolutionary progress in science ofen depends on the use of an
artifcial language. It applies to linguistics, logic, mathematics,
physics and other sciences.
Perhaps the positive argument from Europe was not strong enough
to establish the fourth conclusion. Its negative counterpart in India is
all but decisive. Some of the facts that serve as steps in the argument
are mentioned in Section 4 'Rivers Flowing into the Sea' . In Indian
mathematics, infinite power series were discovered almost three
centuries before it happened in Europe. Indian matematics was, in
this respect, as good as Newton's, but the Sanskrit was not artificial
enough, it was not replaced by equations and the Indian development
ground to a halt.
Indian mathematics
My next conclusion is more tentative. It is elicited by several ques
tions. Indian grocers and other merchants combined the use of place
value (already known to the Babylonians) with the Indian numerals
including the zero that have been called an essential part of the
development of civilization (ogel, 1963: 42).
10
But why did Indian
mathematicians use cumbrous expressions derived from linguistics
9 Regulae ad Directionem Ingenii IV:5 quoted by Gleich (2002: 209) who also refers
to Newton's statement 'The Analysis of the Ancients is more simple more ingenious
& more fit for a Geometer than the Algebra of the Modems'.
10
The idea that the zero orignated in South-East Asia, at the interface of 'the
easter zone of the Indian and the souther zone of the Chinese cultures', espoused
by Needham (1959: 1 0-12) and developed by Lam Lay Young, is not tenable for two
reasons. First, the Indian evidence is earlier (middle of the sixth centur: Hayashi,
2001: 774775). Second, the South-East Asian inscrptions are Indian as is obvious
from Coedes (1931) to which Needham refers. For 'the great movement of inter
course between India and South-east Asia', Needham could, in 1959, only refer to
Grousset and Wales which were about to be outdated by a much more monumental
and conclusive contribution: Coedes (196), itself in due course supplemented by
numerous more recent publications.
ARTIFICIAL LANGUAGES 1 31
(Staal, 1 995a; Yano,hhis volum? Was it to add the prestige of the
/ Qt06J (
science of language to the humdrum activity of mathematical calcu-
lation? Was it because they thought of mathematics itself as a lan-
guage, good for composing verse and telling stories and useful for
doing sums? Was it for excluding outsiders?
1 1
Was it for several or all
of these reasons? I tend to conclude that India's linguistic skill may
have been not merely irrelevant or superfluous, but detrimental to the
development of her mathematics - brilliant as it was.
The generosit of articial language
I now return to the curious fact that equations may tell us something
we did not know or understand before. It does not only apply to
physics but to mathematics itself; and especially to algebra. It was
formulated elegantly by d'Alembert, the leading French mathemati
cian of the mid-eighteenth century (quoted by Boyer and Merzbach,
1991 : 439): 'Algebra is generous; she often gives more than is asked of
her' . Does it tell us something about the world or does it derive from
the ability of language to make infinite use of finite means? Or both?
The mysterious powers of imaginary and complex numbers, which
exhibit that same generosity, were expressed in similar terms by
Leibniz and Roger Penrose. According to Leibniz, these numbers
have an amphibian nature between being and non-being; and yet they
are able to express real quantities and solve real problems (quoted in
Tropfke, 1980, I: 1 53). Penrose (1989: 96) refers to magical properties
that we had no inkling about at first and that are achieving far more
than that for which they were originally designed.
That same magical property was the topic of Eugene Wigner's
1960 lecture The unreasonable efectiveness of mathematics in the
physical sciences. Penrose (1999: 94 and elsewhere) lists it as the fst
mystery of quantum physics. John Barrow's book The Universe that
Discovered Itsel(2000) treats it in a long chapter entitled Why are the
laws of nature mathematical? It contains six pages of an imaginary
dialogue between a Platonist (Penrose's position) and a scientist who
maintains that mathematics is a human invention. The thesis that
begins to bridge the gap between the two positions occurs at the
outset and is the one that is most strongly supported by the history of
11
Not according to the Parahita ('benevolent') system of astronomical compu
tation that was based upon the Aryabhatiya and introduced into Kerala in 683 CE.
A Malayalam commentary explains the name as follows 'since this tradition of
computation was accessible to all people, it acquired the name Parahita' (Sarma,
1954: vi), cf. Mahaevan and Staal (2003 and forthcoming).
132 FRTS STAL
science: We derived most of our mathematics from Nature in the frst
place, so it would be rather surprising i we couldn't then describe
Nature with it.
This thesis does not explain that mathematics is a hua
n
invention,
but supports the idea that it is a human discovery - not the discovery
of a Platonic world separate from the physical universe, but of the
structure of the universe of which we are a part; and the fact that we
are in a position to discover it is due to our innate knowledge of
certain properties of language. How can that be?
Let us fst take a look at an example of how humans derive their
mathematics from nature: the prehistory of the international Theo
rem of PythagorasjBaudhayanajthe Chou Pei Suan Ching. It started
with counting and the natural numbers, represented in many early
civilizations by pebbles, dots, etc. It led to the discovery of triplets
such as (3, 4, 5) with the property that 3 x 3 + 4 x 4 = 5 x 5 (in
modern notation). Measurements of land led to similar findings and
to the idea that the surface of a square with side 3 is a function of 3
(3 x 3 or 3
2
in moder notation). Half squares led to oblongs and
triangles that could be measured with equal ease. Thus emerged the
geometric manifestation of the Theorem leading in due time to many
other new insights and results, all illustrating how discoveries go hand
in hand with notations and other artificial expressions.
The artificial languages of mathematics and physics possess an
inherent knowledge that may remain puzzling even if we get used to it
or interpret it with the aid of intelligible models. Recognition of that
puzzling fact led to a slogan: our goal is inteligibility not of the world
but of theories. Quantum theory put an end to even that. Feynman
(1964 = 1992: 129) famously declared 'Nobody understands quan
tum theory' . Stephen Hawking (1988, 1 996: 232) elaborated: 'Even if
there is only one possible unified theory, it is just a set of rules and
equations' . Such puzzles about meaning are consistent with the fact,
that controversies do not rage about the Schroedinger equation itself
but about the correct interpretation of its meaning (Miller, 2003).
Dirac expressed it as follows: 'My equation is smarter than I am'
(Farmelo, 2003: xvii; Wilczek, 2003).
That intelligence resides in the language of mathematics was noted
by Leibniz. Commenting on the alleged intelligence of mathemati
cians, he observed that when they deal with things other than
mathematics they are not smarter than anyone else. He concluded
that their apparent intelligence is due not to themselves but to the
mathematical language that they have learnt (Mates, 1986: 183). It
ARTIFICI LANGUAGES 133
does not only apply to artifcial language. Every language user knows
that we, whenever we learn a language, learn about the world (not
only the world of humanity). Wilhelm von Humboldt devoted scat
tered portions of the ninth section of the famous Introductin to hs
large work on the Kawi Language of Java to the subject. It is an
obscure piece of writing that endeavors to show, among other things,
how the power of the individual is small compared to the might of
language (wie gering die Kraft des Einzelnen gegen die Macht der
Sprache ist). The topic itself is large but not nebulous and stands in
need of more exploration.
12
Final conclusions
To sum up my own endeavors. My frst conclusion was that artificial
languages exist across sciences and civilizations. It happened inde
pendently in diferent parts of the world. Why? I believe that many of
the facts I have mentioned and several others suggest an explanation.
We know already that the origin of human language caused a speed
up of human evolution. It was the decisive moment that led to
moder humans. My summary conclusion is that artificial languages
are similarly natural and a next step in the same direction. They
appeared in various civilizations and led to scientific, that is, highly
confrmed knowledge. The latest result to date is moder science. It
shows that our tiny planet, endangered as it is by our noxious species,
is a spot where the universe knows itself through language.
REFERENCES
Aoki, Haruo (1983). 'Kana, Kodnsha Encyclopedia ofJapan (pp. 131-135, vol. IV)
Tokyo: Kodansha.
Ariel, Yoav, Shlomo Biderman and Oman Rotem (eds.) (1998). Relativism and
Beyond. Leiden: Brill.
The Aryabhata Group (2000). 'Transmission of the calculus from Kerala to Europe',
Proceedings of the Interational Seminar , and Coloquium on 1500 Years of
Aryabhateeyam (pp. 33-8) Kochi: Kerala Sastra Sahita Parishad.
Barrow, John D. (2000). The Universe that Discovered Itsel. Oxford: Oxford
University Press.
Blay, Michael (1993, 1998). Reasoning with the Infnite. From the Closed World to the
Mathematical Universe. English translation by M.B. DeBevoise, The U
n
iversity of
Chicago: Chicago and London.
12
The idea that language is independent from human interpretations is displayed
through syntax and similar to the idea that ritual is independent from meaning which
I pursued elsewhere (e.g., Staal 1990, 1993).
134 FRTS STA
Bochensk, I.M. (1951). Ancient Formal Logic. Amsterdam: North-Holland
Publishing Company.
Bopp, Franz (181 6).
U
ber ds Conjugationssystem dr Sanskritsprache, in Verglei
chung mit jenem dr griechischen, lateinischen, persischen und germanischen
Sprachen. Frankfurt am Main.
Boyer, Carl B. & Merzbach, Uta C. (1991). A History ofMathematics. New York:
John Wiley & Sons.
Bronkhorst, Johannes (2001a). 'PaJini and Euclid: Refections on Indian geometry'.
Joural ofIndian Philosophy 29, 43-80.
Bronkhorst, Johannes (2001b). 'Pourquoi la philosophie existe-t-elle en Inde?' In
Bronkhorst (ed.): 7-8.
Bronkhorst, Johannes (2001c). 'Indology and rationality'. Asiatische StudinfEtuds
Asiatiqus 55/4, 917-941.
Bronkhorst, Johannes (2002). 'Literacy and rationality in ancient India'. Asitische
Studien/Etuds Asiatiques 56(4), 797-831 .
Bronkhorst, Johannes (e.) (2001). 'La rationalite en Asie/Rationality in Asia. Actes
du colloque de 1 'Institut Interational pour les etudes asiatiques (HAS), Leiden
1999.' Etudes de Lettres, Universite d Lauanne.
Brough, John (1951). 'Theories of general linguistics in the Sanskrit gammarians'
Transactions ofthe Philological Society 27-6. Reprinted in Staal 1972, 402-414
and Brough 1996, 79-98.
Brough, John (1996). In Minoru Hara & J.C. Wright (eds.), Collected Papers.
London: School of Oriental and African Studies, U
n
iversity of London.
Carter, M.G. (1990). 'Arabic gram ar' and 'Arabic lexicography'. In Young et al.
(eds) pp. 106-117.
Chomsky, Noam (2002). On Nature and Language. Cambridge: Cambridge
University Press.
Chomsky, Noam and Halle, Morris (1968). The Sound Patter of Englsh. New
York: Harper & Row.
Coedes, G. (1931). 'Apropos de l'origne des chifres arabes', Buletin ofthe School of
Oriental Studies 6, 323-328.
Coedes, G. (1964). Les hats hindouises d'Indochine et d'Indonesie. Paris: E. de
Boccard.
Cohen, H. Floris (1984). Quantiying Music. The Science ofMusic at the First Stage
ofthe Scientic Revolution: 1580-1650. Dordrecht: D. Reidel.
Cohen, H. Floris (1994). The Scientic Revolution. A Historiographical Inquiry.
Chicago: University of Chicago Press.
Cohen, I. Berard (1983). The Newtonian Revolution. With Illutrations of the
Transformation ofScientic Idas. Cambridge: Cambridge University Press.
Curtius, Erst Robert (1948). Europiische Literatur ud lateiiches Mittelalter.
English translation (1963). New York: Harper & Row.
Danecki, Janusz (1985). Indian phonetical theory and the Arab gram arians.
Rocznik Orientalityczny 4(1), 127-134.
Dani, Ahmad Hassan (1963). Indin Palaeography. Oxford: Clarendon Press.
Daus, Paul (1960). 'Why and how we should correct the mistakes in Euclid'.
Mathematics Teacher 53, 576-581 .
Dijksterhuis, E.J. (1950, 1986). The Mechanization ofthe World Picture. Pythagoras
to Newton. Prnceton: Princeton University Press.
Dyson, Freeman (1988). Infnite in Al Directions. New York: The Author.
Falk, Harry (1993). Schrit i a/ten Indien: Ein Forschungsbericht mit Anmerkungen.
Tibingen: Ginter Narr.
Farmelo, Graham (ed.) (2002a). It Must be Beautiul. Great Equations ofModr
Scince. London and New York: Granta Books.
ARTIFICIAL LANGUAGE 135
Farmelo, Graham (2002b). 'A revolution with no revolutionaries. The Planck
Einstein equation for the energy of a quantum'. In Farmelo (ed.) pp. l-27.
Feynman, Richard P. (1992). The Character of Physical Law. Harmondsworth:
Penguns Books.
Gleich, James (2002). Isaac Newton. London: Fourth Estate.
Grinstead, Eric (1972). Analysis ofthe Tangut Script. Lund: Studentlitteratur.
Gulik, R.H. van (1956, 1980). Siddam: An Essay on the History ofSanskrit Studies
in China and Japan. Nagpur: Satapitaka Series. Vol. 247.
Hacking, lan (ed.) (1981, 1987). Scientic Revolutions. Oxford: Oxford University
Press.
Harbsmeier, Christoph (1998). Language and Logic = Needham, Joseph, Since and
Civilisation in China, Vol. 7, Part I. Cambridge: Cambridge University Press.
Hardy, G.G. (1940, 1994). A Mathematician's Apology. Cambridge: Cambridge
University Press.
Hauser, Marc, Chomsky, Noam and Pitch, W.T. (2002). 'The faculty of language:
what is it, who has it, and how did it evolve?' Scince 298, 22 November.
Hawking, Stephen (1996). The Illustrated A Brief History of Time. Updted and
Expanded. New York: Bantam Books.
Hayashi, Takao (1995). The Bakhshill Manuscript. An Ancient Indian Mathematical
Treatise. Gronigen: Egbert Forsten.
Hayashi, Takao (2001). 'Matematica'. I: Petruccioli (ed.), Ch. X, pp. 772-790.
Haywood, John A. (1960). Arabic Lexicography: Its Hitory, and its Place in the
General History ofLexicography. Leiden: E.J. Brill
Hill, Donald R. (1990). 'Mathematics and applied science'. In Young et al. (eds.)
pp. 248-273.
von Hinuber, Oskar (1990). Der Beginn dr Shrit undfrihe Schritlchkeit i Indin.
Stuttgart: Franz Steiner.
Hodgson, Marshall G.C. (1974). In Edmund Burke Ill (ed.), The Venture ofIslam.
Conscience and History in World History. Chicago and London: The University of
Chicago Press.
Houben, Jan E.M. (1995). The Sa1bandha-Samudsa (Chapter on Relation) and
Bhartrhari's Philosophy ofLanguage. Groningen: Egbert Forsten.
Hoyrup, Jens (1998). On a collection of geometrical riddles and their role in shaping
of four to six Algebras. Filosof og Videnskabsteori pe Roskilde Iniversitetscenter
3/2, 1-68 (Preprint).
Humboldt, Wilhem von (1836).
U
ber die Verschidenheit des menschlichen
Sprachbaues und ihren Eifluss auf di geistige Entwicklung ds
Menschengeschlechts. Berlin: Konigliche Akademie der Wissenschaften. English
translation by Peter Heath (1998). Cambridge: Cambridge University Press.
lngalls, Daniel H.H. (1965). An Anthology ofSanskrit Court Poetry. Cambridge,
MA: Harvard University Press.
Jensen, Hans (1963). Die Schrit in Vergangenheit und Gegenwart. Berlin: Deutsche
Verlag der Wissenschaften.
Kiparsky, Paul (1991). 'Economy and the construction of the
S
ivasutras'. In Madhav
M. Deshpande and Saroja Bhatta (eds.), Pi1inian Studies. Professor S.D. Joshi
Felicitation Volume. University of Michigan: Center for South and Southest
Asian Studies.
Kneale, William and Kneale, Martha (1962). The Development ofLogic. Oxford: The
Claredon Press.
Lam Lay Young and Ang Tian Se (1992). Fleeting Footsteps: Tracing the Conception
ofArithmetic and Algebra i Ancient China. Singapore: World Scientifc.
Law, Vivien (1990). 'Indian Inuence on Early Arabic Phonetics - Or Coincidence'.
Studies in the History ofArabic Grammar I, Proceedings ofthe 2nd Symposiu on
136 FRTS STAL
the History ofArabic Grammar, Nijmegen, 27 April-1 May 1987 (p. 215-227).
Amsterdam/Philadelphia: John Benjamins
Lee, Ki-moon (1983). 'Foundations of Hunmin Chong'um'. In Korean Natinal
Commission for UNESCO (1983), pp. 71-79.
Lw, Bruno (1959). Abriss der japanischen Grammatik. Wiesbaden: Otto Harras
sowitz.
Lloyd, G.E.R. (1990). Demystiying Mentalities. Cambridge: Cambridge University
Press.
Lloyd, G.E.R. (1996a). Adversaries and Authorities. Cambridge: Cambridge
University Press.
Lloyd, G.E.R. (1996b). Aristotelian Explorations. Cambridge: Cambridge Universit
Press.
Lloyd, G.E.R. (2002). The Ambitions of Curiosity. Understanding the World in
Ancint Greece and China. Cambridge: Cambridge University Press.
Lukasiewicz, Jan (1957). Aristotle's Syllogistic. Oxford: The Clarendon Press.
Mahadevan T.P. and Staal, Frits (2003). 'The turg point in a living tradition'.
Electronic Joural ofVedic Studies 10: la (also forthcoming in Indische Kultur Im
Kontext. Festschrit Klaus Myliu, ed. Lars Gohler. Wiesbaden: Otto Harrasso
witz).
Martin, Richard C. and Mark R. Woodward with Di S. Atmaja (1997). Defendrs of
Reaon in Islam. Mu'tazilism from Medieval School to Modr Symbol. Oxford:
Oneworld.
Mates, Benson (1986). The Philosophy ofLeibniz. Metaphysics and Language. New
York and Oxford: Oxford University Press.
Menasce, P. de (1949). 'Note iraniennes'. Journal aiatique 237, 1-6.
Miller, Arthur I. (2003). 'Erotica, Aesthetics and Schrodinger's Wave Euation'. In
Farmelo (ed.), pp. 1 10-131.
Miller, George A. (1964). 'The Psycholinguists'. Encounter 23(1), 29-37.
Miller, Roy Andrew (1967). The Japanese Language. Chicago: The University of
Chicago.
Nakayama, Shigeru and Nathan Sivin (1973). Chinese Science. Exploration ofan
Ancient Tradition. Cambridge.
(
" . {
Narasimha, Roddam (2003 MA.) 'Axiomatism andomputational-rositivism. Two
-athematical-iltures in theirsuit oftheExacciences. Economic and Poltical
Weekly, 36, 50-56 (30 August).
Narasimha, Roddam (forthcoming). Axiomatism and omputatonal "ositivism.
Two Mathematical Cultures in the Pursuit of the Exact Sciences (a Slightly
Extended and Editorially Revised Version of the Same Paper).
Needham, Joseph (1959). Scince and Civilization in China Vol 3: Mathematics and
the Sciences ofthe Heavens and the Earth. Cambridge Cambridge University Press,
Needham, Joseph (1976). Science and Civilzation in China Vol 5: Chemistry and
Chemical Techonology,. Pt.III: Spagyrical Discovery and Invention: Hitorical
Survey, from Cinnabar Elixirs to Synthetic Insulin. Cambridge: Cambridge
University Press.
Nogales, Salvador Gymez (1990). 'Ibn sma'. In Young et al. (eds.) pp. 389-04.
Norman, Jerry (1988). Chinese. Cambrdge: The University Press.
Park, David (1988). The How and the Why. An Essay on the Orgins and Development
ofPhysical Theory. Amherst: University of Amherst Press.
Patzg, Ginther (1963). Die aristotelische Sylogistik. Gottingen: Vandenhoeck &
Rupprecht.
Pa, Octavio (1995). In Light ofIndia (' Vislumbres d la India'). San Diego.
Penrose, Roger (1989). The Emperor's New Mind. Concering Computers, Minds and
the Laws ofPhysics. Oxford: Oxford University Press.
ARTIFICIAL LANGUAGES 1 37
Penrose, Roger with Abner Shmony, Nancy Cartwright and Stephen Hawkng
(1997, 1999). The Large, the Small and the Human Mind. Cambridge: Cambridge
University Press.
Petruccioli, Sandro (e.) (2001). Enciclopedi Italiana: Storia dela Sinza. Vol. 11:
La scienza indina. Coordinaento scentico d Rafaele Torella, Progetto
originario di David Pingree.
Pingee, David (1968). 'The Fragments of the Works ofYa' qub ibn Tiriq'. Joural
ofNear Eater Studies 27, 97-125.
Pingree, David (1970). 'The fragents of the works of al-Fazir'. Joural ofNear
Easter Studis 29, 103-123.
Pingree, David (1970-1994). Census ofthe Exact Sciences in Sanskrit. Series A, Vols.
1-5. Philadelphia: American Philosophical Society.
Pingee, David (1976). 'The Indian and Pseudo-Indian passages in Greek and Latin
astronomical and astrological texts'. Viator 1, 141-195.
Pingree, Davd (1981). Jyotisistra. Astral and Mathematical Literature. Wiesbaden:
Ottto Harrassowitz.
Pingree, David (2003). 'The logic of non-Wester science: mathematical discoveries
in medieval India'. Daedlus 45-53.
Pingee, David (no date). 'Transmission of science' (prepared for the Storia della
Scienza, Vol. II, 2001 , but apparently not published).
Pollock, Sheldon (1985). 'The theory of practice and the practice of theory in Indian
iellectual history'. Joural ofthe American Oriental Society 105: 499.
n
Refou, Louis (1963), 'Sur le genre du sutri dans la litterature sanskrite'. Journal
asiatique 251, 165-216.
Renou, Louis & Filliozat, Jean (1953). L'Ind classique. Manuel ds etuds indinnes,
Vol. II. Paris: Imprimerie Nationale - Hanoi: Ecole franaise d'extreme Orient.
Rundgren, Frithiof (1976). 'Uber den gechischen Einfuss auf die arabische
Nationalgrammatik'. Acta Universitatis Upsaliensi, N.S. 2(5), 1 19-144.
Salomon, Richard (1995). 'On the orign of the early Indian scripts'. Journal ofthe
American Oriental Socity 115, 271-279.
Salomon, Richard (1998). Indian Epigraphy. A Guide to the Study ofInscriptions in
Sanskrit, Prakrit, and the other Indo-Aryan Languages. New York: Oxford
University Press.
Sarma, K.V. (1954). Grahaciranibandana or Parahitagalita by Haridatta. Critically
edited with an Intrduction and Appendix. Mylapore, Madras: Kuppuswami
Sastri Research Institute.
Scharfe, Hartmut (1977). Grammatical Literature. Wiesbaden: Ottto Harrassowitz.
Scharfstin, Ben-Ami (1998). 'A balancing conclusion' In Ariel et al., pp. 301-413.
Scharfstein, Ben-Ami (2001). 'Relativisme et unicite'. In Bronkhorst (ed.), 49-57.
Schwartzberg, Joseph E (1978). A Historical Atlas of South Asia. Chicago: The
University of Chicago Press.
Seeley, Christopher (1991). A History of Writing in Japan. Leiden: E.J. Brill.
Shapin, Steven (1996). The Scientic Revolution. Chicago: Universit of Chicago
de Solla Price, Derek J. (1973). 'Joseph Needham and the science of China'. In
Nakayama and Sivin, (eds.), pp. 921 .
Staal, Frits (1972). A Reader on the Sanskrit Gramarians. Cambridge, MA: The
MIT Press.
Staal, Frits (1975). 'The concept of metalanguage and its Indian background'.
Journal ofIndin Philosophy 3, 31 5-354.
Staal, Frits (1986). 'The fidelity of oral tradition and the origins of science',
Meddlingen dr Koninklike Nederlandse Akadmi van Wetenschappen, Afdeling
Letterkund, n.r. Vol. 49, no. 8, pp. 251-288.
138 FRTS STAL
Staal, Frits (1988). Universals. Studis in Indian Logic and Linguistics. Chicago:
University of Chicago.
Staal, Frits (1989). 'The independence of rationality from literac'. European Journal
ofSociology 30, 301-310.
Staal, Frits (1990, 1993). Rules without Meaning. Ritual, Mantra and the Human
Scinces. New York: Peter Lang.
Staal, Frits (1992). 'Sitra'. In Kapila Vatsyayan (ed.), Kalitattvakosa. A Lexicon of
Fundmental Concepts ofthe Indian Arts. (pp. 355-367) New Delhi: Indira Gandhi
National Centre for the Arts
Staal, Frits (1993-1994). Concepts ofScience in Europe and Asia. Leiden: Intera
tional Institute for Asian Studies.
Staal, Frits (1995a). 'The Sanskrit of science'. Journal of Indian Philosophy 23,
73-127.
Staal, Frits (1995b). 'Concepts of scence in Europe and Asia'. Interdicilinary
Scince Reviews 20(1), 7-19.
Staal, Frits (1998). 'Beyond Relativism'. In Ariel et al. (eds.), pp. 37-66.
Staal, Frits (1999). 'Greek and Vedic geometry'. Joural ofIndian Philosophy 27,
105-127.
Staal, Frits (200la). 'Squares and oblongs in the Veda'. Journal ofIndian Philosophy
29, 257-273.
Staal, Frits (200lb). 'Noam Chomsky between the human and natural sciences', i
Janus Head. Joural ofInterdisciplinary Studies in Literature, Continental Philo
sophy, Phenomenological Psychology, and the Arts, Winter 2001, 25-66. Available
at: http://www.janushead.org./gwu-2001/staal.cfm.
Staal, Frits (200l c). 'Article One'. In Bronkhorst, (ed.) pp. 59-95.
Staal, Frts (2001d). 'Origine e sviluppo dei linguaggi articiali'. In Petruccioli (ed.),
L scienze ne/la cultura Indina, Ch. I, pp. 61 1-38.
Staal, Frits (2003a). 'The Scence of Language'. In Flood, Gavin (e.), The Blackwel
Companion to Hinduism. Oxford: Blackwell.
Staal, Frits (2003b). 'Artical languages: Asian backgrounds or influences?'
Newsletter ofthe Interational Institute for Asian Studis, 3 (March), 15.
Staal, Frits (2004). 'Part of the equation'. The Economit 370/8357 (January 1), 16.
Talmon, Rafael (1997). Arabic Gramar in its Formative Age. Kitib al- 'Ayn ad its
Attriution to Half b. Ahmad. Leiden: E.J. Brill.
Thieme, Paul (1982-1983). 'Meaning and form in the 'Grammar' ofPa:ini'. Studien
zur Indologie und Iranistik 8(9), 3-34.
Tropfke, Johannes (1980). Geschichte dr Elementarmathematik. Band 1: Arithmeti
und Algebra. Vollstandig neu bearbeitet von Kurt Vogel, Katrin Reich und
Helmuth Gericke. Berlin: Waiter de Gruyter.
Truesdell, C. (1968). Essays in the History ofMechanics. New York: Springer.
Verhagen, P.C. (1994). A History ofSanskrit Grammatical Literature in Tbet, Vol.
11. Transmission ofthe Canonical Literature. Leiden: E.J. Brill.
Verhagen, P.C. (2001). A History ofSanskrit Grammatical Literature in Tibet, Vol.
11. Assimilation into Indigenous Scholarship. Leiden: E.J. Brill.
Versteegh, C.H.M. (1993). Arabic Grammar and Qur 'anic Exegesis in Early Islam.
Leiden: E.J. Brill.
Vogel, Kurt (1963). Mohammed ibn Alchwarizmi's Algorismu. Dasfriheste Lehrbuch
zum Rechnen mit indischen Zif er. Aalen: Otto Zeller.
van der Waerden, B.L. (1970). Das heliozentrische System i dr grichischen,
persischen und indischen Astronomie. Zurich: Leemann AG.
Wenek, G. (1954-1959). Japanische Phonetik. Vols. I-IV. Wiesbaden: Otto
Harrassowitz.
ATFCI LANGUAGES 139
Wiger, Eugene (1960). 'The unreasonable efectiveness of mathematics in the
physical sciences'. Communications in Pure & Applid Mathematics 13, 1-14.
Wilczek, Frank (2003). 'A Piece of Magic: the Dirac Equation'. In Farmelo, (ed.),
pp. 1 32-160.
.
Wild, Stefan (1962). 'Neues z a1testen arabischen Lexikographie', Zeitschrit der
Deutschen Morgenlandischen Geselschaft 112(2), 291-298.
Wilkins, Charles (1808). A Grammar ofthe Sanskrita Language. London.
Wujastyk, Dominik (2003). 'The science of medicne'. In Gavin Flood (ed.), The
Blackwel Companin to Hinduism. (pp. 393-9). Oxford: Blackwell.
Yano, Michio (1980). 'Aabhata's Possible Rebuttal to Objections to his Theory of
the Rotation of the Earth'. Historia Scientiarum 19, 101-105.
Yano, Michio (1997). Review of S. Balachandra Rao, India Mathematics and
Astronomy: Some Landmarks (1994). In Galita Bhiratf 19, pp. 128-132.
Young, M.J.L., J.D. Latham and R.B. Sereant (eds.) (1990). Religion, Learing and
Science in the Abbasid Period. Cambridge: Cambridge University Press.
Ziircher, Eric (1959). The Buddhist Conqust ofChia: The Spread and Adaptation of
Buddhism in Early Medieval China. Leiden: E.J. Brill.
http://www.uib.nofpeoplefhnohfjartlang.htm
http:/ fwww.cix.co. uk/-morven/lang/
http:/ fwww.faqsorg/faqs/languagejartifcial-languages-F AQ
International Institute for Asian Studies
Leiden University
P. O. Box 9515
2300 R Leiden
The Netherlands
E-mail: fritsstaal@berkeley.edu