Royal Investigations of the Origins of Language
Robin N. Campbell & Robert Grieve
[from Historiographica Linguistica IX(1/2), 43-75 (1982)]
Students of language development in children are familiar with accounts of three isolation
experiments, that is, experiments in which infants are isolated in order to discover whether
they have a natural tendency to create language and, if so, what sort of language. These
experiments are alleged to have been carried out (in chronological order) by Psamtik I of
Egypt. Frederick II (Hohenstaufen) of Sicily, and James IV of Scotland (cf. Marx, 1967, p.
443, 450-51; Blumenthal, 1970, p. 100).
In view of the present revival of interest in the origins of language, it has seemed to us
timely to examine these ancient investigations from two points of view. First of all, there is
the question of authentication: did they indeed take place and, if so, how were they
conducted and what were their outcomes? In the second place, what light do they shed on
the genetic question to which they were addressed, viz., what is the nature of spontaneously
created language? In the course of our research we stumbled upon a fourth royal
investigation, that of Akbar the Great, 16th-century ruler of Moghul India. This isolation
study is not well-known (see sect. 4 below) and deserves to be better known. In addition,
although accounts of the three earlier experiments are commonplace, they are seldom
accurate and never accompanied by full texts and bibliographic details of source materials.
For these reasons we feel that it is worthwhile to provide contemporary students of child
language with a convenient and comprehensive guide to the origins, weird and wonderful
though they be, of their subject.
We will proceed in the following manner. For each experiment we will give the primary
source-texts (in translation) and refer to any principal secondary texts. After commenting
on these texts we will discuss the question of authenticity. A natural further step is to
consider the plausibility of these investigations. That is, to ask whether it is plausible that
the individual in question should have carried out such a study. As we show below, it turns
out that plausibility is a highly unreliable guide to authenticity. There are two grounds for
this claim. The first is an a priori argument:11 an individual who is known for empirical
inquisitiveness and an interest in genetic or linguistic questions is not only more likely to
carry out isolation experiments than other individuals, he is also more likely to have such
experiments falsely attributed to him by garrulous and imaginative chroniclers (whether
motivated by malice, a desire to ‘gild the lily’ or a simple wish to make their books more
The second ground is empirical: plausibility and authenticity are simply not correlated in
the cases discussed.
Finally, we will consider whether these experiments illuminate the genetic question in
any way and briefly review the possibility of investigating this question by less barbarous
1. The experiment of Psamtik (Psammetichus) I of Egypt
We know of this experiment only through Herodotus’ account2 . Psamtik lived from 663 to
610 B.C., and Herodotus from 485 to 425. His account is based upon what he was told by
the priests of Hephaestus at Memphis. It runs as follows:
‘Now until Psammetichus reigned over them, the Egyptians believed that they were
the eldest of all men. But ever since Psammetichus became king and resolved to learn
who were indeed the eldest, they have believed that the Phrygians were before them,
but they themselves before the rest. For when Psammetichus was not able by enquiring
to learn the answer from any man, he conceived this device. He gave two new-born
babes of ordinary men to a shepherd, to nurture among his flocks after this manner. He
charged him that none should utter any speech before them, but they should live by
themselves in a solitary habitation; and at the due hours the shepherd should bring
goats to them, and give them their fill of milk, and perform the other things needful.
Thus Psammetichus did and commanded because he desired, when the babes should
be past meaningless whimperings, to hear what tongue they would utter first. And
these things came to pass; for after the shepherd had wrought thus for a space of two
years, when he opened the door and entered in, both the babes fell down before him,
and cried becos, and stretched out their hands. Now when the shepherd heard it the
first time, he held his peace; but when this word was oftentimes spoken as he came to
care for them, then he told his lord, and brought the children into his presence when he
commanded. And when Psammetichus had also heard it, he enquired which nation
called anything becos; and enquiring, he found that the Phrygians call bread by this
name. Thus the Egyptians, guided by this sign, confessed that the Phrygians were elder
than they. That so it came to pass I heard of the priests of Hephaestus in Memphis.’
Psamtik is thus thought to have believed that isolated children would speak the language
of their aboriginal ancestors. This belief is presented as an assumption to be used in
interpreting the experimental outcome, not as an hypothesis itself requiring empirical
confirmation. It is characteristic of early writing on the subject of language origins that
some particular language (perhaps ‘dead’ but in some way identifiable) should be regarded
as the original. A particularly clear example of this assumption is provided by the following
passage from the Wibhanga Atuwaba (cf. Hardy, 1966, p. 23), an ancient Singhalese
‘Parents place their children when young either on a cot or a chair, and speak different
things, or perform different actions. Their words are thus distinctly fixed by their
children (on their minds) thinking that such was said by him and such by the other;
and in process of time they learn the entire language. If a child, born of a Damila
mother and an Andhaka father, should hear his mother speak first, he would learn the
Damila language. but if he should hear his father first, he would speak the Andhaka.
If, however, he should not hear either of them, he would speak the Magadhi. If, again,
a person in an uninhabited forest in which no speech is heard, should intuitively
attempt to articulate words, he would speak the very Magadhi.’
Psamtik is credited with having found it necessary to isolate two children (contra the
gedanken-experiments in the Wibhanga Atuwaba), suggesting perhaps that he thought that
language would not be developed without some society. However, it is also possible that
the use of two subjects may have been merely a precaution against inadvertent mortality. In
any event, the outcome of the experiment, as reported, was hardly sufficient to bear the
weight of the inference that ensued. Some have argued that becos might have been
modelled upon the bleating of the goats (cf. Lloyd, 1976, p. 5; Sennert, 1643, sect.VI;
Farrar, 1865, p. 13), but this seems equally tendentious and futile. It is interesting that
Psamtik is supposed to have asked in what tongue is there anything called becos rather than
in what tongue is there something called becos which would fit the occasion of utterance.
What would he have said if his advisers had informed him that becos was Egyptian for the
tusk of an elephant? Could it be that semantic emptiness or lability in the Original
Language was considered possible but that phonological change was not? At any rate it is
clear that a latent Ursprache (perhaps consisting only of a collection of meaningless
phonological forms) was assumed to be present in the minds of the children by Psamtik (or
by the fabricators of the tale, if it was indeed only legendary).
Herodotus’ report of this experiment has several features that might incline one to accept
it as genuine. It bristles with impressive detail, the design shows remarkable sophistication
(secure isolation, newborn children, minimal caregiving, awareness of babbling, etc.) and
the outcome was presumably unwelcome to the Egyptians (they knew perfectly well who
were the oldest people!). However, accepting it as genuine is quite unjustified. A recent
commentary on Book II (Lloyd, 1976) makes this very clear.
Firstly, the sophisticated nature of the design suggests an Ionian origin --Greek doctors
were already employing such controlled methods in Ionia (cf. Farrington, 1944; 1969).
Secondly, goats’ milk was not used for suckling infants in Egypt (indeed, its only known
application was as a remedy for anal dysfunction) but was so used in Ionia (cows’ milk was
not). Thirdly, there is a remark, at the end of the account, to the effect that the story was
known to Greek scholars in a slightly different form. Lloyd argues convincingly that a
version of this experiment had appeared earlier in Hecateus of Miletus’ Genealogies (only
fragments of this survive) and it is possible that Herodotus may have simply ‘lifted’ the
story from there (for exhaustive discussion see Heidel, 1935).
There are still various possible alternative explanations. For example, it might have been
a fictional satire aimed at foolish Egyptians or it might have actually happened in Ionia and
somehow ‘migrated’ to Egypt. However, whatever its provenance, any claim that the
experiment took place in this way or in that way is seriously weakened by this analysis of
Lloyd’s, since, at best, it greatly increases the remoteness of Herodotus’ report from the
Herodotus describes an experiment of Psamtik’s which might be thought of as an
investigation of a genetic question, namely an attempt to fathom the depth of the springs
then regarded as the source of the Nile, but since this is no better authenticated than the
language experiment it does not help much. In the case of Psamtik, then, we know neither
that such an experiment was consistent nor that it was inconsistent with his personal
characteristics, since so little is known about him.
2. The Experiment of Frederick Il (Hohenstaufen) of Sicily
Once again, we know of this experiment through only one source, the chronicle of
Brother Salimbene, a Franciscan friar. The story occurs in the Chronicle as one of a long
list of strange experiments recounted by the friar and runs as follows:
‘Like Psammetichus, in Herodotus, he made linguistic experiments on the vile bodies
of hapless infants, bidding foster-mothers and nurses to suckle and bathe and wash the
children, but in no wise to prattle or speak with them; for he would have learnt
whether they would speak the Hebrew language (which had been the first), or Greek,
or Latin, or Arabic, or perchance the tongue of their parents of whom they had been
born. But he laboured in vain, for the children could not live without clappings of the
hands, and gestures, and gladness of countenance, and blandishments.’3
Although, unlike the previous case, Salimbene (1221-88) was a contemporary of
Frederick (1194-1250), he had no connection with his court, and his personal
knowledge of Frederick was limited to a glimpse of him passing through the streets of
Parma in 1235. It would be unwise, therefore, to suppose that Salimbene’s sources of
information were reliable. This is particularly the case since Frederick, though Holy
Roman Emperor and active in Crusades, was excommunicated by Pope Gregory IX in
1227 and thereafter lived in almost continual conflict with the Church. It is therefore
likely that Salimbene would have included in his Chronicle any discreditable tale that
reached his ears (or his imagination).
As for the text itself, once again we have as a background assumption that some
identifiable language should be the basis of the Ursprache. However, in this case, as
well as the usual classical languages, it is thought possible that the language of the
parents might be produced, in this case probably a vernacular, Sicilian. At this time the
classical languages and the vernaculars were regarded as being quite different, only the
former involving ‘necessary’ forms. There is an illuminating contemporary discussion
of this point by Dante,4 who argues that the outcome of experiments such as this
would inevitably favour the classical languages.
The explanation of the outcome of Frederick’s experiment is equally striking. In
Western educational thought it is not until the early 18th century that such weight
begins to be placed on the social and affective aspects of early childrearing. As we
shall show below, Frederick was an extremely original thinker who cared nothing for
received opinion and whose grasp of the elements of scientific investigation was not to
be equalled for several centuries. These unusual aspects of Salimbene’s report add to
its credibility, since they are unlikely inventions.
Frederick was possibly the most unusual monarch in European history. In his time
he was called stupor mundi et immutator mirabilis, ‘wondrous transformer of the
world’, and with good reason. Two excellent biographies, by Kantorowicz (1928) and
by Van Cleve (1972), and the specialist investigations of Haskins (1924, 1929) are the
principal sources for the sketch which follows. Frederick’s strongest personal
intellectual interests were biological. He assembled an immense menagerie which
accompanied him whenever he travelled. It included camels, lynxes, leopards, lions,
panthers, apes, bears, a giraffe and an elephant; hawks, owls, eagles, buzzards, falcons,
peacocks, parrots and an ostrich. In addition, his court was always accompanied by a
collection of human curiosities: concubines, slave-girls, eunuchs, acrobats, conjurors,
rope-dancers, etc. He maintained animal reservations in various parts of his kingdom,
the most remarkable of which was a collection of waterbirds housed in ‘natural’ (i.e.,
not in a zoo) conditions in a marsh near Foggia. Like many mediaeval rulers he was
passionately fond of falconry but here again his interest was scientific as well, so
much so that he wrote a treatise, De Arte Venandi cum Avibus, which is the basis of all
subsequent writing on the subject.5 This treatise is a large work and it contains an
introduction of about 100 pages dealing with avian biology in general and describing
many simple investigations of anatomy and behaviour carried out by Frederick.
Aristotle is taken to task on more than one occasion for mistakes in his natural
histories. Frederick is never in any doubt which should be given greater weight, the
wisdom of the great philosophers and teachers or the evidence of observation and
experiment. Of course, in other quarters, such as the mediaeval universities, such
empirical questioning of received knowledge would have been regarded as obscene
and heretical. There is little doubt that Frederick carried out other sorts of experiments
as well. In the list provided by Salimbene some are undeniably dubious, but others
have some substantiation - such as the ‘ringing’ experiment to discover the longevity
of fish (cf. Hauber, 1912) - and others, like the isolation experiment, are sufficiently
unusual in design or conception to make us doubt whether they are fabrications. For
example, to compare the effects of activity and inactivity on digestion he had two men
eat the same meal, sleep and go hunting respectively, whereupon they were
simultaneously disembowelled and their stomach contents examined!
Of Frederick’s other intellectual interests we should mention at this point his
interests in language. He is reputed to have spoken nine languages and to have been
literate in seven. He anticipated the vernacular poetry of Dante by composing many
Sicilian odes and lyrics and, most remarkably, inspired the officials of his court to
write vernacular poetry as well! It is clear that the imputed isolation experiments is
entirely consistent with what is known about Frederick’s interests and resources.
3. The Experiment of James IV of Scotland
This experiment is reported in the History of Robert Lindesay of Pitscottie as
‘The king gart tak ane dum woman and pat hir in Inchekeytht and gaif hir tua zoung
bairnes in companie witht hir and gart furnische them of all necessar thingis pertening
to their nurischment that is to say, meit, drink, fyre and candell, claithis, witht all
wther kynds of necessaris quhilk (is) requyrit to man or woman desyrand the effect
hierof to come to knaw quhat langage thir bairnes walk speik quhene they come to
lauchful aige. Sum sayis they spak goode hebrew bot as to myself I knaw not bot be
the authoris reherse. (Thir actis foirsaid was done in the zeir of god Im iiijc lxxxxiij
Like the previous case, Pitscottie (1500?-1565) had no personal acquaintance with
the monarch, nor personal knowledge of the experiment, since James died in 1513.
‘Lauchful aige’ is ‘lawful age’. Exactly what is meant by this is unclear: it may have
been that age at which children were thought capable of sin, that is, according to the
Church, 7, or it may have been the age at which children ceased to be ‘pupils’ - 12 for
girls, 14 for boys.7 Pitscottie’s final remark is clearly a signal to the reader that the
reported outcome is to be regarded as a humorous invention, since his entire history up
to that point and for a decade or so beyond is based upon ‘authoris reherse’ - hearsay.
The experiment is, unlike the previous two, characterised as open-ended, no particular
possible outcomes being specified. It is plainly a less satisfactory account evidentially,
than either of the previous texts. Herodotus described Psamtik’s experiment in an
impressively detailed way and Salimbene’s account, as we have noted, had certain
features which make it an unlikely fabrication. But Pitscottie’s lacks details of
procedure and outcome and, indeed, has all the characteristics one would expect of a
colourful fabrication - an uninhabited island as the venue (i.e. no-one could bear
witness that it had not taken place) and a ludicrous outcome. It does, however, have
two features that the others lack, a date and a place. The island of Inchkeith lies in the
Firth of Forth, readily accessible from James’ palaces in Edinburgh and Linlithgow. It
was certainly visited by James during the 1490s for the purposes of falconry and he
maintained a mews there (cf. Treasurer’s Accounts8 ). However, there are no entries in
the Accounts which relate to the experiment (such as, e.g., payments to the dumb
woman). Moreover, Inchkeith was used from 1497 onwards as a repository for the
victims of venereal disease, introduced to Edinburgh by the followers of Perkin
Warbeck - a pretender to the English throne.9 It is a small island and it seems likely
that the experiment would have had to be concluded before this use for purposes of
quarantine. Of course, it could be that the date given, 1493, refers to the conclusion of
the experiment. But if this is so, then the experiment is even less credible since James
was only 20 in 1493. On the whole, then, we have no very good grounds for taking
Pitscottie’s story as genuine. Despite this, we have made very extensive inquiries in
search of additional historical evidence, but without any substantial results. There are
no records surviving from either the Parish of Kinghorn (in which Inchkeith is located)
or the Abbey of Inchcolm, a neighbouring island in the Firth. As footnote 9 shows, the
surviving state papers are not complete, but contain no mention of the experiment.
Neither do the unofficial contemporary writings of De Ayala10 or the court poet
Dunbar (Mackenzie, 1932). As well as that of Pitscottie there are three other 16th-
century histories, Buchanan (ed. Ruddiman, 1715), Drummond of Hawthornden
(1655), and Lesley (1830). None of them mentions the experiment. The solitary ray of
hope that there may be more to the story than frivolous invention comes from
Pinkerton’s 18th-century history (1797, p. 25) in which there is an account of the
experiment (later embellished by Thomson,1893), which differs from that given by
Pitscottie. Pinkerton was a normally reliable professional historian and we can be sure
that he must have had good reason for giving a different account. But what was his
reason? He cites, in addition to Pitscottie, a location in Buchanan (Chap.XIII, sect.7)
but in Ruddiman’s edition of Buchanan there is no mention of the experiment, at that
place or at any other. Pinkerton’s account runs as follows:
‘To make some discovery on the origin of language, two infants under the charge of a
dumb woman were sent into the isle of Inchkeith: but the self-taught speech has not
been explained: and it is needless to add that it must have been original, and perhaps
though there was some society, little superior to the brutish babble of those
unfortunate beings, lost during infancy in extensive forests.’
So, according to this account, there was (undeciphered) spontaneous speech. Pinkerton’s
own observations - ‘it is needless . . .’ show that he was acquainted with the great debate
amongst continental philosophers and educators which took place at this time and which
had as one of its foci the nature of feral children (cf. Lane, 1977, for a thorough account of
this). It seems to us a distinct possibility that somewhere in that massive literature there is
material relating to this and other isolation experiments but so far we have been unable to
find it. It may be, then, that the last word has not been said about this putative experiment
of James. However, it must be conceded that, on available evidence, it seems unlikely that
the experiment was performed.
There is a mass of evidence which shows that James was interested in experiment and in
the development of the sciences, particularly medicine. Moreover, like Frederick he was a
collector of curiosities. To take the last point first, he gave instructions to his merchant
seamen (all of whom engaged in occasional piracy) that he would reward them for any
human or animal oddity (cf. Chambers and Thomson, 1855). It is also fairly certain that he
kept and maintained a joined twin. There are accounts of this in Pitscottie (Vol.I, p. 233)
and in Drummond of Hawthornden (1655, p. 69). We shall give the account which appears
in Aikman’s translations from Buchanan’s Latin (for purposes of clarity):
‘About this time a strange kind of monster was born in Scotland. In the lower part of
the body it resembled a male child, differing in nothing from the ordinary shape of the
human body, but above the navel, the trunk, and all the other members became double,
and were distinct, both in their use and appearance. They caused it to be carefully
brought up, and educated, particularly in music, in which it wonderfully excelled. It
also learned different languages, and in their various inclinations, the two bodies
appeared to disagree between themselves, sometimes disputing, each preferring
different objects, and sometimes consulting, as if for the common pleasure of both,
and, what was remarkable, when the lower limbs, or loins were struck, both felt the
blow in common, but when pricked or otherwise hurt above, only one of the bodies
was sensible of pain, which distinction was most conspicuous in death; for, when the
one body had died several days before the other, that which remained, when the dead
half became putrescent, wasted away by degrees. I write this the more confidently,
because there are many persons of undoubted veracity still alive, who saw the
Buchanan’s dating of the monster’s birth is 1490 and Pitscottie states that it lived for 28
years and, during this time was maintained by the court. (For a more recent case, see
Gedda, 1951.) James’ interests in medicine and science have been amply substantiated by
Guthrie (1497) and Read (1938a, 1938b & 1947, Chap. 2). His new university at Aberdeen
had a separate medical faculty (albeit one devoted to the customary doctrinaire and
theoretical methods of teaching) and he founded the more practically-minded college of
surgeons and barbers in Edinburgh, whose graduates had to satisfy their examiners that they
‘knew anotamell [anatomy], nature and complexion of every member humanis bodie, and . .
. the vaynis of the samyn, that he may mak flewbothomell [phlebotomy] in dew tyme . . .
(for) every man aucht to knaw the nature and the substance of every thing that he werkis, or
ellis he is negligent’. Moreover, each member was entitled to ‘anis in yeir ane condempnit
man efter he be deid to mak anatomell of, quhairthrow we may haif experience’ (cited in
James himself was an enthusiastic surgeon: the Treasurer’s Accounts show payments to
individuals who had their teeth extracted by him, and suchlike. In science, James founded
what Read (1938b) has described as Scotland’s first research laboratory in Stirling Castle
around 1500. The director of this establishment was a mysterious Italian, John Damian de
Falcusis (later appointed Abbot of Tongland - a sinecure) who came to Scotland around that
time from Paris. Damian was an alchemist who became a close friend of James. He was
provided with excellent facilities for his work, as the Treasurer’s Accounts show, in
addition to a position at the Court and luxurious garments of office (cf. Read, 1938b). There
is no evidence either that he succeeded in multiplying his seed gold or that he was ever
seriously called to account. However, another of Damian’s experiments apparently was
subjected to public test - human flight! The experiment is described in William Dunbar’s
hilarious and savage satirical poem, The Fenyeit Frier of Tungland (ed. by Mackenzie,
1932) and also by John Lesley, Bishop of Ross (see Lesley,1830, p. 76).
Like Frederick, we can reasonably suppose that James had an interest in language and
poetry. According to the Spanish ambassador to the Scottish court, Don Pedro de Ayala,
James was a wonderful linguist, speaking Latin, French, German, Flemish, Italian, Spanish
and Gaelic, but Mackie (1958, p.118) regards this as a diplomatic embellishment. It is
certain, however, that James fostered vernacular poetry, which flourished as never before
(nor since) during his reign. Dunbar, whom we have already mentioned, was the finest of
these poets and was attached to James’ court.
It should now be evident that it is wholly consistent with what is known about James that
he should have carried out the isolation experiment. However, the direct evidence, as we
have noted, is very weak.
4. The Experiment of Akbar the Great, Moghul Emperor of India, 1542-1605
Akbar’s experiment is by far the most interesting of these royal investigations. In the first
place, it is the only one that is adequately vouched by contemporary documents. There are
three primary sources for the experiment: Abu’l-Fazl, Badauni, and Xavier (see below) -
and a large number of secondary sources some of which, though not contemporary, are still
of interest. Some of these are referred to below. In the second place, Akbar’s experiment
was superior to the other putative experiments in conception and design. Lastly, Akbar’s
experiment has received hardly any attention from psychologists or linguists this century. It
is described as a matter of course, along with the other experiments in some 19th-century
texts, e.g., Farrar (1865), Muller (1861, pp. 480-82) and Tylor (1878, pp.79-81), but after
Tylor does not appear to have been mentioned in linguistic or psychological texts until
Panconcelli-Calzia (1937, 1955: reference only to Catrou, 1705 - a poor secondary source)
and Borst (1957-63 IV, p. 2050: reference to the same source). Marx (1967) and
Blumenthal (1970), though including surveys of early investigations, fail to mention
Akbar’s experiment. More recently, Hewes (1977, p. 98) and Blakemore (1976) have
mentioned it, though once again without adequate detail or references. We ‘discovered’ it
by reading Panconcelli-Calzia (op.cit.) as a result of reading Marx, who evidently had not
read it himself.11 Muller, as well as being a prominent linguist, specialized in Indian history
and hence was thoroughly familiar with the Akbar experiment. However, he made little
reference to it in his copious writings, possibly because of his rather exotic views on the
subject, to wit,
‘It is useless to inquire whether infants, left to themselves, would invent a language. It
would be impossible, unnatural and illegal to try to experiment, and without repeated
experiments, the assertion of those who believe and those who disbelieve the
possibility of children inventing a language of their own are equally valueless!’ (1861,
Thus Muller takes the possibility of replication as an essential criterion of ‘genuine’
experiments and, presumably, empirical observations in general. For a student of history
this is a fairly extraordinary point of view!
We shall begin by giving some relevant details of Akbar’s life and character, since we
assume that neither Akbar himself, nor the mode of life of a Moghul Emperor, is familiar to
most readers of these pages.
The standard biographical source for Akbar for many years has been Smith (1917), which
contains a useful, but biased bibliography of contemporary material. Smith, like most
biographers, concentrates on economic and political aspects of Akbar’s reign and, in fact,
fails to mention the experiment. Smith’s biography has now been superseded by Srivastava
(1962, 1967), which is much superior. Good sources for the religious and intellectual
history of the period are MacLagan (1932) and Aziz (1969).
Akbar became Moghul Emperor on the death of his father in 1556. His early career
showed no particular unusual features, except for a notable degree of religious tolerance
(for instance he abolished a poll-tax on non-Muslims and took non-Muslim wives) which
may well have been adopted as a political expedient, and a taste for exotic sports, such as
pigeon-flying and polo (played in the dark with slow-burning balls of lignite) as well as the
traditional royal pursuits of falconry and hunting. Hunting of large game in Moghul India
was a particularly bloody business, involving the slow encroachment over several days of a
60 mile diameter circle of beaters on the hunters located at the centre. In one such
qamargha during Akbar’s reign 15,000 animals were slaughtered.
The first sign of eccentricity appears in 1575, when Akbar built the ‘Ibadat-Khana (Hall
of Worship) in his palace at Fatehpur Sikri, just south of Agra. This was, despite its name, a
debating hall, built in the form of a cross. Akbar was illiterate and, naturally, greatly valued
oral debate, although he possessed, and was said to be thoroughly acqainted with, a library
of 24,000 books and manuscripts. The four wings of the cross were occupied during debates
by the Ularna (Islamic jurists and professors), the Sayyids (descendents of Muhammad), the
Shaikhs (wizards and prophets) and finally, dilettante members of Akbar’s court. Some
details of the subjects of debate are provided in The Dabistan12 and Badauni13. It is 1578,
however, which marks the intellectual watershed of Akbar’s life. At this time he
experienced a kind of vision and, as a result, adopted a thoroughly free-thinking, though
still spiritualised, mode of life. He introduced various non-Muslim regimes, e.g., special
fasts, into his personal habits, abandoned the murderous qamarghas (described above) and
introduced into the debates in the ‘Ibadat-Khana Christians, Hindus, Jain Buddhists, Jews,
Zoroastrians and Sabeans. He established himself in partial authority over the faith by
means of the so-called Infallibility Decree of 1579 and invited the first of three Jesuit
missions from Goa to his court. These missions failed in their object of conversion, Akbar
having deep rooted objection to (1) obedience and submission to the Church and (2)
dismissal of all but one wife, the latter action being, of course, politically impossible. In
fact, it is clear that Akbar’s interest in Christianity was almost entirely academic. He wished
to introduce diversity into the debates in the ‘Ibadat-Khana and also to obtain fresh ideas
about possible faiths. In 1582 he promulgated, with the connivance of Shaikh Abu’l Fazl,
his closest advisor, a new faith - the Din-i-Ilahi - which was eclectic and ecumenistic in
nature, bringing together what Akbar perceived as being the most valid doctrines of the
religions that he had studied. This faith did not survive him. The language experiment,
which we now describe, was carried out during the period 1578-82 and thus coincides with
this great intellectual and spiritual crisis in Akbar’s life.
We give below, in full, the three contemporary accounts of the experiment. The first of
these is from the Akbarnama of Abu’l-Fazl.14 The Shaikh was a close personal confidant
and advisor and, like Akbar himself, a partisan of free-thinking attitudes to Islam. Although
Abu’l-Fazl’s history is given a low valuation by Smith (1917) and also by other Europeans
such as Beveridge (cf. footnote 16) and Haig (1937), this is called in question by Srivastava
(1962) and by Aneer (1973). The work is often described as tedious, repetitious, full of
pointless flattery and obscure allusions. That may be, but nothing factual in the Akbarnama
has been faulted and the work was based closely upon state papers and other court records
including aide-memoires recounting Akbar’s every action and word. Akbar (like other
Moghul emperors) kept a close eye upon posterity and employed scribes to follow him and
compile these aide-memoires for the benefit of his historian, Abu’l-Fazl. For these reasons
we feel that the greatest weight should be given to his account of the experiment, which
‘One of the occurrences was the testing of the silent of speech. There was a great
meeting, and every kind of enlightenment was discussed. In the 24th Divine year
(1578) H.M. said that speech came to every tribe from hearing, and that each
remembered from another from the beginning of existence. If they arranged that
human speech did not reach them, they certainly would not have the power of speech.
If the fountain of speech bubbled over in one of them, he would regard this as Divine
speech, and accept it as such. As some who heard this appeared to deny it, he, in order
to convince them, had a serai built in a place which civilized sounds dit not reach. The
newly born were put into that place of experience, and honest and active guards were
put over them. For a time tongue-tied wet-nurses were admitted there. As they had
closed the door of speech, the place was commonly called the Gang Mahal (the dumb-
house). On the 9th August 1582 he went out to hunt. That night he stayed in Faizabad,
and next day he went with a few special attendants to the house of experiment. No cry
came from that house of silence, nor was any speech hear there. In spite of their four
years they had no part of the talisman of speech, and nothing came out except the
noise of the dumb. What the wise Sovereign had understood several years before was
on this day impressed on the hearts of the formalists and the superficial. This became a
source of instruction to crowds of men. H.M. said, ‘Though my words were proved,
they still are saying the same things with a tongueless tongue. The world is a miserable
abode of sceptics. To shut the lips is really to indulge in garrulity. They have
hamstrung the camel of the Why and Wherefore, and have closed the gate of speech
with iron walls.
Enough, Nizami, be silent of discourse,
Why speak to a world with cotton in its ears,
Shut your demonstrations into a narrow phial,
Put them all in a phial and place a stone thereon.’
We may note, for further reference, two points of ambiguity. First, it is not clear whether
the speech, whatsoever it might be, is to be regarded as Divine or whether the type of
speech that it is (e.g., Persian, Hebrew, etc.) is to be counted as Divine. That is we do not
know whether Akbar entertained two hypotheses (no speech, some particular known
speech) or three (no speech, some unknown speech, some particular known speech).
Second, something ‘came out’, viz. ‘the noise of the dumb’. It is unclear to us exactly what
is meant by this. Beveridge makes no additional comment here. Although a scholar of 16th
century Persian might be able to make this more precise, it seems likely that the outcome
will remain essentially ambiguous: were they simply ‘noises’ or perhaps a kind of language,
The second contemporary account is by Badauni. Badauni was an orthodox Sunni
Muslim and detested Akbar’s religious vagaries. His history contains much derogation of
these tendencies and, accordingly, was not published until after Akbar’s death. Srivastava
(1962) regards it as bigoted and less reliable than the Akbarnama and observes that the
translation of the relevant volume by Lowe is inaccurate.
‘At this time they brought a man to Court, who had no ears nor any trace of the
orifices of the ear. In spite of this he heard everything that was said to him, though the
place of the ears was quite level. And in this year, in order to verify the circumstances
of this case, an order was issued that several suckling infants should be kept in a
secluded place far from habitations, where they should not hear a word spoken. Well-
disciplined nurses were to be placed over them, who were to refrain from giving them
any instructions in speaking, so as to test the accuracy of the tradition which says:
‘Every one that is born is born with a natural tendency’, by ascertaining what religion
and sect these infants would incline to and above all what creed they would repeat. To
carry out this order about twenty sucklings were taken from their mothers, for a
consideration in money, and were placed in an empty house, which got the name of
‘Dumb-house’. After three or four years they all turned out dumb and the appellation
of the place turned out prophetic. Many of these sucklings became the nurselings of
My mother is earth, and I am a suckling,
The propensity of children for their mother is strange,
Soon will it be that resting from trouble
I shall fall drunk with sleep on my mother’s bosom.’
We may note that the connection between the earless man and the experiment is, to say
the least, obscure! The ‘natural tendency’ referred to is, of course, to Islam. The final
metaphor is clearly for death.
The third and last truly contemporary account is by Hieronymus Xavier, who had charge
of the third Jesuit mission to Akbar in 1598. The account appears in a letter of Xavier’s,
which was published in a number of different locations15 . We give the translation by
Beveridge (1888). A convenient source for this translation is MacLagan (1932). Xavier is
reporting a conversation with Akbar himself:
‘He told me that nearly 20 years ago he had 30 children shut up before they could
speak, and put guards over them so that the nurses might not teach them their
language. His object was to see what language they would talk when they grew older.
He was resolved to follow the laws and customs of the country whose language was
that spoken by the children. But his endeavours were a failure, for none of the children
came to speak distinctly. Wherefore at this time he allowed no law but his own.’
Although there are some discrepancies with the account in the Akbarnama, the details are
very similar and there is a perfect correspondence of dates. Xavier, like Badauni, attributes
a purely religious motive to Akbar. Abu’1-Fazl is, as we have seen, much less clear about
this. After all, according to Abu’1-Fazl, Akbar’s own belief (in advance of the experiment)
was that there would be no power of speech. It is interesting that the motive, even if
religious, is still different from the motives of Psammetichus and James IV, which were
Turning now to secondary sources, we have managed to discover five that are not far
removed in time from the date of the experiment. These are Purchas (1626, p. 516), Sennert
(1643), the Dabistan (footnote l3), Manucci (1653ff.) and Catrou (1705). Purchas’ and
Sennert’s accounts are clearly based on Xavier’s and add nothing to it. The others differ, in
different ways, from the primary accounts and are given in full.
From the Dabistan (ca. 1648):
‘In like manner, a number of children were put in a place called Gangmahel, where
everything necessary was furnished to them; but none could articulate a letter; having
remained there to their fourteenth year, they were found to be dumb, which made it
evident, that letters and language are not natural to man, that is, cannot be used unless
they have been acquired by instruction, and it is then only that the use of conversation
becomes possible. From this the conclusion was drawn that the world is very ancient
and language of long date.’
From Manucci (1653ff.):
‘Akbar had been anxious for a long time to satisfy two subjects of curiosity, which he
kept in his heart. The frst was to know what language a child would speak who had not
the use of speech or any master to teach it. The second was to find the source of the
famous river Ganges. For the first of these inquiries he ordered the erection of a house
with many rooms at a distance of six leagues from the city of Agrah, and directed
them to place in it twelve children, who should be retained there till the age of twelve
years. An injunction was laid on everyone that, under pain of death, no one should
speak a word to them or allow them to communicate with each other. This was done,
because one set of men asserted that they would speak the natural language, that which
was the language of our first parents. Others held that they would speak the Hebrew
language; others that they would not speak anything but Chaldean; while the Hindu
philosophas and mathamaticians asserted that they must infallibly speak the Samscript
(Sanskrit) language, which is their Latin. However, the twelve years having passed,
they produced the twelve children before the king. Interpreters for the various
languages were called in to help. Each one put questions to the children, and they
answered just nothing at all. On the contrary, they were timid, frightened, and fearful,
and such they continued to be for the rest of their lives.’
From Catrou (1705):
‘It may be said, that curiosity and a thirst for knowledge were the ruling passions of
Akebar. His indulgence of these propensities prompted him to a very singular
experiment. He was desirous to ascertain the language in which children would
express themselves, who had been kept in ignorance of the articulate sounds of any
known language. The emperor had been informed, that the Hebrew was the original
language of the human race, and the one, which all, who had not been taught any
other, would naturally speak. In order to secure a conviction on this point, he ordered
twelve children to be taken from the breast, and to be closely confined in a castle,
which was situated six leagues from Agra. They had given to them, for nurses, twelve
women, who were dumb, with the addition of a man, who was also dumb, to save as
porter. The porter was forbidden, on pain of death, ever to open the gates of the castle.
When the children had attained the age of twelve years, Akebar commanded that they
should be brought into his presence. He then assembled in his palace persons skilled in
various languages. A Jew, who was at Agra, was appointed to the office of deciding,
whether the language to which they might give utterance, was Hebrew. The capital
furnished Arabians and Chaldeans in abundance. The Indian philosophers, on their
side, contended, that the children would speak the Sanscrit, which is the dialect of the
learned of the country, and holds among them the same place, as does the Latin among
the learned in Europe. The ancient books of philosophy and the Indian theology are
written in this language. When these children appeared before the emperor, to the
surprise of every one, they were found incapable of expressing themselves in any
language, or even of uttering any articulate sounds. They had learnt, from the example
of their nurses, to substitute signs for articulate sounds. They used only certain
gestures to express their thoughts, and these were all the means which they possessed
of conveying their ideas, or a sense of their wants. They were, indeed, so extremely
shy, and, at the same time, of an aspect and manners so uncouth and uncultivated, that
it required great labour and perseverance to bring them under any discipline, and to
enable them to acquire the proper use of their tongues, of which they had previously
almost entirely denied themselves the exercise.’
These secondary accounts are of dubious value. To varying extents they differ from the
contemporary accounts and, naturally, from each other. This is true even of Catrou and
Manucci, and in that case we know that the former is directly based upon the latter. In the
case of discrepancy we must (other things being equal) prefer the contemporary accounts.
So we may reasonably discount the claim in Manucci that the children were not allowed to
communicate with one another, which makes a nonsense of the whole experiment. Why
undertake the cost and trouble of isolating a large group if isolating one child would be
sufficient? The Dabistan is interesting on account of its remarkably perceptive conclusion.
Catrou’s is undoubtedly the most provocative account, since it comes closest to what we
would suppose to be the most likely outcome of such an experiment (cf. discussion below).
However, without arduous and costly bibliographic research (which would very likely be
fruitless) it is impossible to evaluate Catrou’s claims about the outcome of the experiment.
As is commonplace in historical research, the state papers and other fundamental
documents available to Catrou are now lost (Srivastava, 1962).
At present, then, we are justified (or so it seems to us) in assuming that Akbar’s
experiment tells us only that a system of communication based upon speech cannot be
created in a single generation. This is the strongest claim compatible with Akbar’s data. It is
possible that, had the period of isolation been longer, such a system would have been
developed. Equally, it is possible that there was a rudimentary system of this sort but that its
presence went unnoticed. The difficulty of distinguishing language from gibberish (cf.
glossolalia) is well-known.
We would like to begin by drawing attention to the scientific isolation of this experiment
of Akbar’s. As far as we know there is no general treatment of the development of scientific
thought in India (as there is, for example, in the case of China, cf. Needham, 1954- ).
Islamic influence on the development of astronomy, chemistry and medicine in the West
has been considerable, and Frederick II’s anachronistic empiricism can perhaps be traced to
the Islamic faction within his court. Yet, a recent survey of the intellectual influence of
Islam in India (Aziz, 1969) fails to mention any scientific involvement.
Although in the case of Akbar it is superfluous to investigate the plausibility of his
experiment, since the direct evidence is so strong, it is worth doing so since it shows that
plausibility (obviously not a sufficient condition for attributing such an experiment) is not a
necessary condition either. There is no convincing evidence that empirical procedures were
common at the time of Akbar in India. So Akbar’s experiment must be regarded as
inconsistent with what other information we have about his background and scientific
interests. Indeed, the only grounds on which we might claim that the experiment was a
plausible one are that Akbar, as shown by his activities in the ‘Ibadat Khana, had a strong
interest in fundamental religious questions, many of which are genetic in nature. But the
tremendous costs of his fantastic many-sided debating forum surely suggest that Akbar
regarded scholarly argument, rather than experiment, as the proper way of resolving such
questions! Accordingly, we ought to conclude that it was not in character for Akbar to carry
out this experiment. And yet he surely did, so it may be that, although it seems to be an
entirely natural procedure, our practice of assessing plausibility may be quite valueless as
far as the question of authenticity is concerned.
5. General Discussion
Plainly there is little to be gleaned from these experiments which bears on the genetic
question. The only experiment with a relevant outcome is Akbar’s and there, as we have
already remarked, the outcome is ambiguous. The ‘noise of the dumb’ reported by Abu’l
Fazl and corroborated by Xavier could have been practically anything from silence to fully-
fledged language. Moreover, in the absence of speech we would expect there to have been
some gestural communication (see below for substantiation) and yet this was not reported
in the contemporary accounts.
The second issue to be dealt with here is more troublesome: can we, accepting the
impossibility of contemporary ‘royal investigations’, do anything to answer the genetic
question today? Our discussion of this issue will be brief and rather dense
First of all, it seems to us that the classical question, ‘What language will be spoken by
children isolated from birth?’ could and should be made more precise. A number of
relatively free parameters can be recognised, as follows:
(1) the presence of one or more caretakers;
(2) the number of children isolated;
(3) the ages of the children at the start of the experiment;
(4) the period of isolation;
(5) the set of communicative resources available to each ordered pair of individuals
(channel) in the group. By this is meant the repertoires of detectable (i.e., visible,
audible, tactile) gestures able to serve as signs;
(6) the system for communication that is (by virtue of experience and/or heredity)
natural to each individual;
(7) the system for communication which has been established (by intra-group contact)
for each ordered pair of individuals (channel) in the group;
(8) the communicative load during the period of isolation--that is, the interests, wishes,
intentions, needs, etc., of the ordered pairs of members of the group insofar as they
(a) are novel and (b) relate to joint action and interaction, and
(9) the physical circumstances of isolation.
For any particular experiment, some of these parameters will be constant ((1), (2), (3),
(4), the natural system of the caretakers, and (9)) during the period of isolation. The
remainder will vary with ongoing sensory and motor development (5), affective and
cognitive development (8) or with all parameters (the natural systems of the children, and
(7)). When different possible experiments are considered, however, all of the constant
parameters must be regarded as variables. In addition, some of the variable parameters may
be varied between experiments: most obviously, selection of deaf and/or dumb caretakers
and/or children will interfere radically with parameter (5). Although parameter (6) has a
clear interpretation for caretakers and will in general be a natural language, it is less clear
for the case of the children. What we intend here is the minimal system of meaningful
audible and visible gesture which is given by inheritance together with (for points later in
the course of an experiment or for later starting points) the modification and augmentation
of this system which results from development of the systems of variable (7). Roughly, for
each child, the system mentioned in (6) will be revised so as to correspond with the union
of the systems of variable (7) for each channel involving that child. Now, since the values
of all parameters other than (7) are either (a) available for direct manipulation or (b)
dependent on relatively autonomous psychological development, we may think of the array
of values mentioned in (7) as the output of a very general time-function of the other eight
The classical question may then be seen as a question about the value of this function (let
us call it the language acquisition function, LAF) for a limiting16 set of values of its
arguments, namely where no resources are available on any caretaker-originating channel
and where no system of communication has been established on any child-originating
channel. In fact, it is evident that this is not only a limiting case of the general function but
also an unattainably ideal one. In practice, no caretaker could be ‘pruned’ of communicative
resources and still be able to take care. Equally, children who were isolated without a
caretaker before their communicative resources had developed to the point where systems
of communication could be established would simply not survive. It is therefore quite plain
that the classical question (construed as this ideal case) cannot be answered directly, i.e.,
empirically. However, if we could discover the nature of the general function LAF (by
determining its value for other empirically and ethically feasible values of its arguments)
then we could give a theoretical answer to the classical question. In fact, we know a little
about LAF already. For example, Feldman et al. (1977) have shown that in the case of only
children who are profoundly deaf and who are raised by parents who deliberately inhibit
conventional visible gesture, the output of LAF after 3 years or so is a system of visible
gesture substantially more complex than the residue of natural gesture (pointing, etc.)
employed by these caretakers. Further, it has often been noted (cf., e.g., Stokoe, 1974;
Tervoort, 1961) that groups of young deaf children will develop (with minimal assistance
or, indeed, despite active opposition) quite complicated gestural systems well-adapted to
their communicative load. So it seems as if LAF has some limited powers of augmentation,
a position which is also implicit in the hypothesis that the ‘creole continuum’ cf. Bickerton,
1974) is the synchronic consequence of rapid development of an original pidgin by
successive generations (Hall, 1966).
Another kind of ‘natural’ experiment which bears on LAF is the case of twins. It is well
known that a high proportion of twins (identical or otherwise - develop ‘secret languages’
for communication within the twin (cf. Zazzo, 1960 II, chap.4). Although these are often
marginal and simple this is not always the case. In addition to the more or less inscrutable
cases cited by Hale (1888) and Jesperson (1922) there is the spectacular modern case of
Poto and Cabengo (Orange, 1977)17 . Whether such cases provide evidence for the position
that LAF is an augmentative function is unclear but it seems certain at least that quite
radical alterations are possible. Finally, there is a natural source of evidence which has only
recently begun to be tapped, viz., cases where children with normal hearing are reared by
deaf parents. Potentially, such cases are promising since they come so close to the
conditions of the classical isolation experiment. However, modern urban cases are
complicated by the factors of television and hearing relatives, both readily available
speaking caretakers. Some systematic work has recently been reported by Schiff and Ventry
(1976) and Schiff (1979). The latter paper gives a comprehensive bibliography of previous,
unsystematic case reports. While such children typically show eventual deficits which may
persist through middle childhood (Schiff and Ventry, 1976), they swiftly surpass the oral
language of their principal caretakers and, at any rate in the first year and a half of language
acquisition, their oral language develops in much the same fashion as normal-caretaker
controls. In Schiff’s words (1979, p. 581) “children, when cognitively ready, need little
exposure to the normal model language to learn to speak during the early stages of
development”. Certainly, there is nothing in this research which speaks against the notion
that LAF is augmentative and much that speaks for it.
However much information such natural experiments may yield, it remains unlikely that
it will be sufficient to determine the function LAF. Moreover, the timespan of such
investigations is perforce long and their moral character (given the alternative of
intervention) dubious at best. A more practical, if less gaudy, method for obtaining further
information about LAF would involve parents’ voluntarily adhering to speciific
communicative regimes for short periods during development whilst closely monitoring the
systems for communication employed by the children during this period. Natural
experiments which approach this characterisation have now been quite widely investigated
(cf. the studies collected in Snow and Ferguson, 1977). All that would additionally be
involved is the element of control over caretaker-originating channels and control or
measurement of the communicative load.
As a final point, we would like to mention the old argument due to Hale (1886, 1888) that
the geographical distribution of unrelated language isolates is correlated with the
probability of abandoned groups of children managing to survive. Thus, in hostile
environments, such as the Arctic regions or the great deserts, no unrelated languages are to
be found, while in regions where the climate is favourable and food is plentiful (Hale’s
favourite example was California) many unrelated languages are found. Hale used this
presumed correlation as evidence supporting his view that abandoned groups of children
would routinely re-invent language. This now seems rather hopeless, since the number of
apparently unrelatable languages has shrunk so dramatically since Hale’s time. The
question which now arises is whether the lack of language isolates in favourable areas
constitutes evidence against Hale’s view that rapid language invention is a routine
consequence of isolation? That is, granting the possibility of rapid language re-invention
under conditions of isolation, should there not be many more unrelated languages than there
appear to be? We think not: the processes of pidginisation and creolisation show how
massive and rapid the effects of contact can be. It seems more reasonable to suppose that
unrelated language isolates are a consequence of geographical, political or economic
isolation, rather than of natural isolation experiments!18
Aneer, Gudmar. 1973. Akbar the Great Mogul and his Religious Thoughts. Uppsala: Skriv
Armen, Jean-Claude. 1971. Gazelle-boy: A child brought up by gazelles in the Sahara.
Translated by Stephen Hardman. London: Bodley Head.
Aziz, Ahmad. 1969. An Intellectual History of Islam in India. Edinburgh: Edinburgh Univ.
Bergenroth, Gustav Adolph (1813-65), ed. 1862-68. Calendar of Letters, Despatches and
State Papers relating to the Negotiations between England and Spain. Presented in the
Archives at Simancas and elsewhere. 2 vols. & supplement. London: Longmans, Green.
Beveridge, Henry (1837-1929). 1888. Father Jerome Xavier. Journal of the Asiatic Society
of Bengal 57:1-38.
Bickerton, Derek. 1975. Dynamics of a Creole Continuum. London: Cambridge Univ.
Blakemore, Colin. 1970. A Burning Fire. The Listener 96:2487.743-44.
Blumenthal, Arthur L. 1976. Language and Psychology: Historical aspects of
psycholinguistics. New York: J. Wiley.
Borst, Arno. 1957-63. Der Turmbau von Babel. 4 vols. Stuttgart: A. Hiersemann.
Brown, Peter Hume (1849-1918).1891. Early Travellers in Scotland. Edinburgh: David
Brown, Roger W. 1958. Words and Things. Glencoe, Illinois: Free Press.
Buchanan, George (1506-82). 1715. "Rerum scoticarium historia". Opera Omnia. Ed. by
Thomas Ruddiman ( 1674-1757). Edinburgh: Robt. Freebairn.
Buchanan, George (1506-82). 1827. The History of Scotland, translated from the Latin.
With notes and a continuation to the Union in the reign of Oueen Anne. Translated by
James Aikman (1779-1860). Glasgow: Blackie, Fullarton.
Catrou, Francois (1659-1737). 1705. Histoire generale de l'Empire du Mogol depuis sa
fondation. Sur les memoires portugais de M. Manouchi, Venitien. Paris.
Catrou, Francois (1659-1737). 1709. The General History of the Mogol Empire, from its
foundation by Tamerlane to the late Emperor Orangzeb. (Translator unknown). London:
Chambers, Robert, and Thomas Thomson, comps. 1870. A Biographical Dictionary of
Eminent Scotsmen. Glasgow: Blackie & Son. [Vol. 1. 96; entry for Andrew Barton.]
van Cleve, Thomas Curtis. 1972. The Emperor Frederick II of Hohenstaufen: Immutator
mundi. Oxford: Clarendon Press.
Coulton, George G. 1907. From St. Francis to Dante. London: David Nutt.
Curtiss, Susan. 1977. Cenie: A psycholinguistic study of a modern-day 'wild child'. New
York: Academic Press.
Drummond of Hawthornden, William (1585-1649). 1655. The History of Scotland from the
year 1423 until the year 1842, etc. London: H. Hills.
Farrar, Frederick (1831-1903). 1865. Chapters on Language. London: Longmans, Green.
Farrington, Benjamin. 1936. Science in Antiquity. Oxford: Oxford Univ. Press.
Farrington, Benjamin. 1944. Greek Science; its meaning for us. Harmondsworth, England:
Feldman, Heidi, Susan Goldin-Meadow, and Lila Gleitman. 1977. Beyond Herodotus: The
creation of language by linguistically deprived deaf children. Action, Gesture and
Symbol: The emergence of language ed. by Andrew Lock, 351-414. New York:
Gedda, Luigi. 1951. Studie dei gemelli. Rome: Edizioni Orizzonte Medico.
Guthrie, Douglas. 1947. King James IV of Scotland: His influence on medicine and
science. Bulletin for the History of Medicine 21.173-92.
Haig, Lt.Col. Sir Thomas Wolseley (1865-1938). 1937. Akbar. The Cambridge History of
India. Vol.4 (The Mughul Period), Chapters 4&5, pp. 70-155. Ed. by Sir Wolseley Haig
& Sir Richard Burn. London: Cambridge Univ. Press.
Hale, Horatio Emmons (1817-96).1886. The Origin of Languages. Proceedings of the
American Association for the Advancement of Science 35: 279-323.
Hale, Horatio Emmons (1817-96). 1888. The Development of Language. Proceedings of
the Canadian Institute [ 3rd series] 6: 1-44.
Hall, Robert A., Jr. 1966. Pidgin and Creole Languages. Ithaca, N.Y.: Cornell Univ. Press.
Hardy, Robert Spence (1803-68). 1866. Legends and Theories of the Buddhists, compared
with history and science. Edinburgh: Williams & Norgate.
Haskins, Charles Homer (1870-1937). 1921. The 'De Arte cum Avibus' of the Emperor
Frederick II. English Historical Review 36:334-55.
Haskins, Charles Homer (1870-1937). 1924. Studies in the History of Mediaeval Science.
Cambridge, Mass.: Harvard Univ. Press.
Haskins, Charles Homer (1870-1937). 1929. Studies in Mediaeval Culture. Oxford:
Hauber, Anton. 1912. Kaiser Friedrich II der Staufer und der langlebige Fisch. Archiv fur
die Geschichte der Naturwissenschaften und der Technik 3: 315-29.
Heidel, William Arthur (1868-1941). 1935. Hecataeus and the Egyptian priests in
Herodotus, Book II. Memoirs of the American Academy of Arts and Sciences [New
Series] 18(2): 53-134.
Heizer, Robert Fleming and Albert B. Elsasser. 1963. Original Accounts of the Lone
Woman of San Nicolas Island. Aboriginal California: Three Studies in Culture History
ed. by R. F. Heizer, 125-81. Berkeley, Calif.: Univ. Calif. Archeol. Research Facility.
Herve, Georges (1855-1932). 1909. Le premier programme de l'anthropologie. Bulletins et
Memoires de la Societe d 'Anthropologie [Serie V] 10:473-87.
Hewes, Gordon Winant. 1975. Language Origins. 2 vols. The Hague: Mouton.
Hewes, Gordon Winant. 1977. A Model for Language Evolution. Sign Language Studies 1
Hewes, Gordon Winant. 1978. Review of J.H. Stam's Inquiries into the Origin of Language
(New York: Harper & Row, 1976). Historiographica Linguistica 5:174-89.
Jespersen, Otto (1860-1943). 1922. Language, its Nature, Development and Origin.
London: Allen & Unwin.
Kantorowicz, Ernst Hartwig. 1928. Kaiser Friedrich der Zweite. Berlin: C. Bondi.
Kantorowicz, Ernst Hartwig. 1928. Frederick the Second, 1194-1250. Translated by Emily
O. Lorimer. London: Constable.
Lane, Harlan. 1977. The Wild Boy of Aveyron. Cambridge, Mass.: Harvard Univ. Press.
Lesley, John [Bishop of Ross] (1527-96). 1830. The History of Scotland. Edinburgh:
Lloyd, Alan B. 1976. Herodotus Book Il: Commentary 1-98. Leiden: E.J. Brill.
Mackenzie, William Mackay. 1932. The Poems of William Dunbar. Edinburgh: Porpoise
Mackie, Robert Laird. 1958. King James IV of Scotland. Edinburgh: Oliver & Boyd.
MacLagan, Sir Edward Douglas. 1932. The Jesuits and the Great Mogul. London: Burns,
Oates & Washbourne.
MacLean, Charles. 1977. The Wolf Children. London: Allen Lane.
Manucci, Niccolo (17th cent.) 1907-08. Storia do Mogor: or, Mogul lndia 1653-1708, by
Niccolo Manucci, Venetian. Translated with introduction and notes by William Irvine.
London: John Murray for the Royal Asiatic Society's Indian Texts Series.
Marwick, Sir James David (1826-1908), et al. 1869-82. Extracts from the Records of the
Burgh of Edinburgh. Edinburgh: Scottish Burgh Records Society.
Marwick, Sir James David (1826-1908), et al. 1871. Charters and other Documents relating
to the City of Edinburgh. Edinburgh: Scottish Burgh Records Society.
Marx, Otto. 1967. The History of Biological Basis of Language. Biological Foundations of
Language by Eric Lenneberg, Appendix B. New York: J. Wiley.
Muller, Friedrich Max (1823-1900). 1861-64, Lectures on the Science of Language. 2 vols.
London: Longmans. (Quoted after the 1891 reprint.)
Needham, Noel Joseph Terence Montgomery. 1954- . Science and Civilisation in China.
London: Cambridge Univ. Press.
Nicholson, Ranald. 1974. Scotland: The Later Middle Ages. Edinburgh: Oliver & Boyd.
Ogburn, William Fielding, and Nirmal K. Bose 1959. On the Trial of the Wolf Children.
Genetic Psychology Monographs 60: 117-93.
Orange, J. 1977. The Strange Case of Poto and Cabengo. Woman (17 December), 38-41.
Panconcelli-Calzia, Guilio (1878-1966). 1937. Der 'wilde Knabe' in seiner Beziehung zu
manchen Sprachstorungen im Kindesalter und zu der Entstehung von Mythen.
Medizinische Welt 11: 410-14.
Panconcelli-Calzia, Guilio (1878-1966). 1955. Das Motiv vom 'Wilden Knaben'. Zur
Sprache verwilderter Kinder. Sprachforum 1: 272-77.
Pinkerton, John (1758-1826). 1797. The History of Scotland from the Accession of the
House of Stuart to that of Mary. 2 vols. London: C. Dilly.
Powell, John Enoch. 1949. Herodotus' Histories. 2 vols. Oxford: Clarendon Press.
Purchas, Samuel the Elder (1577?-1626). 1626. Purchas his Pilgrimage. 4th edition.
London: W. Stansby.
Read, John. 1938a. Alchemy in Scotland. The Chemist and Druggist (25 June), 14.
Read, John. 1938b. Alchemy under James IV of Scotland. Ambix 2(2): 60-67.
Read, John. 1947. Humour and Humanism in Chemistry. London: G. Bell.
Schiff, Naomi B. 1979. The Influence of Deviant Matemal Input on the Development of
Language during the Pre-School Years. J. of Speech and Hearing Research 22(3): 581-
Schiff, Naomi B., and Ira M. Ventry. 1976. Communication Problems in Hearing Children
of Deaf Parents. J. of Speech and Hearing Disorders 41: 348-58.
Sennert, Daniel (1572-1637). 1643. Cur, qui surdi sunt a nativitate, idem et muti sint?.
Paralipomena Book 2, Part 3(2). Leiden.
Singh, Joseph Amrito Lal, and Robert Mowbry Zingg. 1942. Wolf-Children and Feral Man.
New York: Harper.
Smith, Vincent Arthur (1848-1920). 1917. Akbar the Great Mogul, 1541-1605. Oxford:
Snow, Catherine E., and Charles A. Ferguson. 1977. Talking to Children. London:
Cambridge Univ. Press.
Srivastava, Ashirbadi Lal. 1962-67. Akbar the Great. 2 vols. Agra, India: Shiva Lal
Agarwala & Co.
Stokoe, William C. 1974. Classification and Description of Sign Languages. Current
Trends in Linguistics 12: 345-72. The Hague: Mouton.
Tervoort, Bernard Th. 1961. Esoteric Symbolism in the Communicative Behaviour of
Young Deaf Children. American Annals of the Deaf 106(5):436-80.
Thomson, Thomas. 1893. A History of the Scottish People from the Earliest Times. 3 vols.
Glasgow: Blackie & Son.
Tylor, Sir Edward Burnett (1832-1917). 1878. Researches into the Early History of
Mankind. 3rd ed. London: John Murray.
Wood, Casey Albert (1856-1942), and Florence Majorie Fyfe. 1943. The Art of Falconry:
The 'De arte venandi cum avibus' of Frederick II. Stanford, Calif.: Stanford Univ. Press.
Zazzo, Rene. 1960. Les jumeaux, le couple et la personne. 2 vols. Paris: Bibliotheque
Zingg, Robert Mowbry. 1940. Feral Man and Extreme Cases of Social Isolation. American
Journal of Psychology 53: 487-517.
1 We are indebted to Professor G. Barrow of St. Andrews University for making this point
clear to us.
2 Herodotus’ well known Histories are of course available in a number of editions with a
range of alternative translations and commentaries. The text presented below is from the
English translation by Powell (1949:109-10) of Book II, section 2.
3 The text given is from Coulton’s (1907) translation of part of the Chronicle. An edition of
the original chronicle, Cronica Fratis Salimbene de Adam Ordinus Minorum, appears in
Monumenta Germaniae Historiae, XXXII, Scriptores, ed. by O. Holger-Egger, p. 350.
4 In De Vulgare Eloquentia, translated into English, with notes, by A.C. Ferrers Howell.
(London: Kegan Paul, Trench, Trubner & Co. London, 1890).
5 Bibliographical details and a review from the standpoint of the history of science can be
found in Haskins (1921). A modern edition in English with copious illustration and
accompanying background material is available (Wood and Fyfe, 1948).
6 Robert Lindesay of Pitscottie’s Historie and Cronicles of Scotland exists in several
manuscripts. The (1899-1931) edition by A.J.G. Mackay (Edinburgh: Wm. Blackwood
for the Scottish Text Society) is based on two of the oldest MSS. It is well-known as a
colourful and unreliable history (cf. the remarks in Nicholson 1974:328, 627). The text
reproduced below is from volume 1, p. 237 of Mackay’s edition.
7 This latter suggestion was offered by Professor G. Donaldson of Edinburgh University.
8 (Treasurer’s Accounts) Compota Thesaurariorum Regum Scotorum: Accounts of the Lord
High Treasurer of Scotland, ed. by T. Dickson and Sir J. Balfour Paul (1877-1916).
These accounts, though incomplete for certain years (e.g., 1499), together with the
Exchequer Rolls of Scotland, ed. by J. Stuart and others (1878-1908) provide a fair
record of the official expenditure of the Crown. As such they are perhaps the most
important primary documents for the royal history of the period. Indeed, there are few
other reliable records from the mediaeval period in Scotland that survive.
9 Mentioned in Edinburgh Burgh records (cf. Marwick 1869-82, 1871). Inchkeith was used
in 1475 for a similar purpose - quarantine of plague victims (cf. Exchequer Rolls, VIII,
364). Apart from these incidents, Inchkeith’s principal function was as an important
10 Ambassador to Scotland from the Court of Spain. His letters are collected in volume 1 of
Bergenroth (1862-68), and a translation of the material relating to James IV can be found
in P. Hume Brown ,1891.
11 One may reasonably speculate that other investigations may come to light. For example,
there is a tantalizing reference by M. Robert Réboul in Hervé (1909) to an attempt to
carry out an isolation experiment by Louis-François Jauffret in the early l9th century.
Jauffret was the founder of the famous Société des Observateurs de l’Homme. Hewes
(1978) contains further details of this possible ‘lost’ experiment. Although we have
described this method of investigation as ‘barbarous’, this last example shows that the
method’s appeal is not limited to the mediaeval mind. Indeed, at the time of writing this
note we have received news of an attempt by a Californian group to obtain funds (and
presumably permission) for such an experiment in the remote southern seas . .
12 The Dabistan is a Persian manuscript thought to have been written by one Muhsin Fani
around 1648. It deals with various religions and, in particular, with the origins of
Akbar’s own faith, the Din-i-Ilahi The text given below is extracted from section Vlll of
David Shea’s English translation (Paris, 1843).
13 Mulla Abdul Qadir Badauni published a three-volume history in Persian (the second
volume being devoted to the reign of Akbar) shortly after Akbar’s death, the
Muntakhab-ut-Tawarikh. The whole work has now been translated into English and
published in the Bibiliotheca Indica series by the Asiatic Society of Bengal. The text
given below is from the translation of volume II by W. H. Lowe (1884), p. 296
14 The Akbarnama, once again a Persian manuscript, was translated into English by H.
Beveridge and published in 3 volumes by the Asiatic Society of Bengal in their
Bibiotheca Indica series (1897-1910). The text which follows appears on pp. 581-2, sect.
393, volume III of that edition.
15 Xavier’s letter (in Latin) originally appeared in a pamphlet by John Oranus of Liege
(1601), Japonica, Sinensia, Mogorana, hoc est, de rebus apud eas gentes a Patribus
Societas Jesu anno 1598 et 99 gestis. The original Latin and an English translation
appear in Beveridge (1888).
16 We shall not discuss the most extreme limiting case, where a single infant is isolated. As
is well-known, there have been a number of celebrated ‘natural’ experiments of this sort
(cf. for reviews, Panconcelli-Calzia 1935 & 1937; Zingg 1940, and Brown 1958
(Chap.V); for the most interesting such cases, see Lane 1977, Curtiss 1977, and the
strange and possibly fanciful Armen 1971). It is sufficient to note that in no such case
has the isolate shown signs of being able to communicate in a systematic way and to
record our own view that it is inconceivable that any should, except for the sort of
communication system that might be established with an animal caretaker. From the
point of view of this paper, a more interesting ‘natural’ experiment would be one in
which two or more infants were accidentally isolated. However, apart from the dubious
case of the wolf-children of Midnapore (Singh and Zingg 1942; Ogburn and Bose 1959;
MacLean 1977), who showed no reported signs of ability to communicate, there are no
usable reports of such cases.
17 According to Edward Klima (personal communication), the ‘secret language’ of Poto
and Cabengo appears to be a version of English.
18 An additional ground for not discarding Hale’s hypothesis is that it is highly likely that
such natural isolation experiments have occurred and will continue to occur in politically
turbulent areas such as Central Africa and SouthEast Asia. The displacement of the
Californian Indian peoples resulted in one known case (the Lone woman of San Nicolas
Island, see Heizer and Elsasser 1963) which might well have been a case of the type
Hale envisaged (but was not). Also, Moffat (1842, p. 10-11) in a discussion of the
varieties of dialect found among the Balala and Sechuana Bushmen of Southern Africa
observes that ‘[on hunting and scavenging expeditions] fathers and mothers and all who
can bear a burden, often set out for weeks at a time, and leave their children to the care
of two or more infirm old people. The infant progeny, some of whom are beginning to
lisp, while others can just master a whole sentence, become habituated to a language of
their own .... In the course of a generation the entire character of the language is