by Keith Rawlings, April 2003

Updated March 2011

A Note to my Readers Concerning this Update:

The information on my main web pages is focused on Asian puppet forms for the most part. Except for the shadow play below, I have written these essays to include some additional European data, plus some of my new findings and theories, the result of further studies. I have not added a bibliography for these articles, as I mention the titles, dates, and authors in the body of the text. However, should it be required to know the exact sources of any of the information here or anywhere on my site, please do not hesitate to contact me at:


Home Page

More on the Ancients
Early Evidence of the “Swazzle”?
The Real Origins of Puppets and Other Thoughts
The Origins of Shadow Theatre
The Earliest Records of Shadow Plays in Europe


It may be odd to consider, but there is much evidence to indicate that the modern marionette was preceded by a more complex device, namely the neurospasta, worked by cords running through and inside the figure, pulled by the demonstrator, either from below, above, or to the side. Consider the following ancient texts, quoted from Michael Byrom’s The Puppet Theatre in Antiquity, 1996, with the words for “puppet” re-translated to a more literal meaning:

“...the outside is moved from within like a [moving figurine, sigillario].” Quintus Septimus Tertullian(us), De Anima, Chap 6, Sec 3, c. 155 - 220 AD.

“Slavery cannot make you do anything, there is nothing outside that can draw the strings within.” Aulus Persius Flaccus,  Saturae, 5, c. 34 - 62 AD.

“This is the device, I think, that is used in moving [images] with cords; for passing over their articulations, the cords are fastened to the beginning of the parts beyond, so that the [images]  readily obey the force of the upward pull when the cords are tightened.” Claudius Galen(us), De usu partium, Book 1, Chap 17.

“For just as those who move [images] with cords attach them beyond the joints to the heads of the members to be moved, so Nature long ago used the same device at all the joints.”  Ibidem, Book 3, Chap 16, c. 129 - 199 AD.

It is generally true that simpler forms are usually followed by more complex ones, but here the opposite seems to have occurred. There is more documentary evidence of complex self-moving figures (automata), and at an earlier date, than manually manipulated puppet figures. (Ancient Egypt and Greece used mobile statues in their religious cults - moved by any number of devices.) And this is true not only in ancient Greece and Rome, but also in ancient China and India. To understand how this came about, we can view this as follows.

The earliest mention of neurospasta is by Herodotus describing an Egyptian practice, and the procedure of these in Egypt was that of a wooden figure held aloft on a long pole, which figure had a phallus presumably worked by a string hidden inside the pole, pulled from below. Many historians have mentioned that such figures were known in ancient Rome, but until recently I had not located the original texts. I had recourse before this only to puppet historians who do not supply the precise references. I quote their conclusions here: “...colossal animated effigies which were used in the victory parades and processions that preceded the circus games at Rome. Among these was the Manducus, a human monster with horrific jaws and teeth which could be made to clash terrifyingly together by means of hidden cord” (Byrom, 1996). “In Rome during the religious ceremonies preceding games in the circus wooden statues were carried which moved their heads and pretended to attack one another” (Joseph Spencer Kennard in his Masks and Marionettes, 1935). Now finally, I have located the original references, through Magnin’s famous book, Histoire des marionnettes en Europe depuis l’antiquité jusqu’à nos jours, second edition, 1862. The most detailed reference is in Pompeius Festus’ De Verborum Significatu, under Manduci. This work is a Latin lexicon, written in the second century AD. My rough translation of this entry follows (I have used an English quote of Plautus from another source):

Manduci: effigies with great, gaping jaws, used by the ancients in processions, making noise with their teeth, inciting laughter in others and causing fright, of which Plautus says [in Rudens (The Rope), act II Scene, VI, verse 51]: ‘How about hiring myself out at some fair as a wild man [manduco]’ ‘Why that?’ ‘Because of the grand way I gnash my teeth.’

Also enlightening in this regard is a passage from Rabelais’ Gargantua and Pantagruel, Book 3, Chap 31, written in 1546 AD:


In this order they moved [in a procession] towards Master Gaster, after a plump, young, lusty, gorbellied fellow, who on a long staff fairly gilt, carried a wooden statue, grossly carved, and as scurvily daubed over with paint; such a one as Plautus, Juvenal, and Pomp. Festus describe it. At Lyons [France] during the Carnival it is called Maschecroute or Gnawcrust; they call’d this Manduce.

It was a monstrous, ridiculous, hideous figure, fit to fright little children; its eyes were bigger than its belly, and its head larger than all the rest of its body; well mouth-cloven however, having a good pair of wide, broad jaws, lined with two rows of teeth, upper tier and under tier, which, by the magic of a small twine hid in the hollow part of the golden staff, were made to clash, clatter, and rattle dreadfully one against another; as they do at Metz with St. Clement’s dragon.

His mention of the name Manduce, obviously derived from Manducus, shows perhaps that ancient Roman “puppetry” and theatre survived in the amusements of the common man up to and through the Middle Ages. Whether or not drama and theatre died out completely after the fall of Rome is a much-disputed point among theatre (and puppet) historians. On the side of those who do not believe theatre continued underground after Roman times, is the fact that Rabelais himself was familiar with classical literature, as seen by his frequent references to the classical authors in his fictional writings. He may be describing real activities of his contemporaries, i.e., the Manduce procession, or these may be all fictional, influenced by his reading of the classics (he does indeed refer to those very classical authors who are the source of our limited information on Manducus).

Next, we have mention of private entertainments (see Xenophon, in my Chapter One) using neurospasta. The practice of showing these is mentioned as laughable, so either some kind of humor was involved, or perhaps the figures just looked and moved comically. Plato, in his Laws (Book II) mentions what is normally translated as “puppets” given as an entertainment that was shown to children. Again, I have replaced the translations as “puppets” with the literal meanings:

ATH. [Athenian Stranger]: …suppose a man were to organize a competition, without qualifying or limiting it to gymnastic, musical or equestrian sports…and, proclaiming that this is purely a pleasure-contest in which anyone who chooses may compete, should offer a prize to the competitor who gives the greatest amusement to the spectators, - without any restrictions as to the methods employed…what do we suppose would be the effect of such a proclamation?

CLIN. [Cleinias]: In what respect do you mean?

ATH.: The natural result would be that one man would, like Homer, show up a rhapsody, another a harp-song, one a tragedy, and another a comedy; nor should we be surprised if someone were even to fancy that he had the best chance of winning with a [wonder-show, thaumaston]. So where such as these and thousands of others enter the competition, can we say who will deserve to win the prize?…do you wish me to supply you with the answer to this absurd question?

CLIN.: By all means.

ATH.: If the tiniest children are to be the judges, they will award the prize to the showman of [wonders, thaumata] will they not?

The Athenian Stranger goes on to discuss the other entertainers and which other kinds of audiences would award these other entertainers the prize.

From this text, we can extract that the showman (a conjuror or acrobat, jongleur) performed alone, and that it was an entertainment popular with children. But the term thaumata in this text is very ambiguous. As no mention here of figures or strings accompanies the word, thaumata could be referring to magic tricks or acrobatics - the root meaning of thauma is “wonder, marvel.” But elsewhere, Plato does use this word in connection with figures, namely in his Republic, Book VII (see my Chapter One for a discussion of the reference to thaumata in the Republic, and my Chapter Two for a more extensive section of the same text). In addition, Plato, in his Laws, Book I, uses this word in connection with string-controlled figures:

Let us imagine that each one of us is a [wonder, thauma] created at the hands of the Gods for their own amusement…The impulses that motivate us are like so many strings pulling in different directions.

Following this, we have notices of more complex movements - the neck, the eyes, the arms, etc. The Roman physician Galen notes for us that these figures were internally strung, and says so in very precise language (see above), so that there is no doubt about it. Whether or not these figures performed plays and farces, and whether or not they spoke any dialogue, is another much disputed point.

The questions confronting us are:

* Did these puppets speak? - implying that they performed plays

*  Were they merely mechanical dancing figures (automata)?

* Are these ancient puppets the “fathers” of our modern puppets? In other words, did our modern puppets descend directly from the ancient Greek and Roman?

The above questions summarize the main issues raised by puppet historians Michael Byrom in England (in his The Puppet Theatre in Antiquity) and Hans Richard Purschke in Germany (in his Die Anfänge der Puppenspielformen und ihre vermutlichen Ursprünge, 1979) concerning the puppets of antiquity. (I have already touched upon the question of direct unbroken descent from ancient Rome for puppets, above. As for the other questions I will attempt to present some evidence for discussion.)

* Byrom believes that there was a puppet theatre and farces performed using puppets in ancient Greece and Rome

* Purschke believes there were no plays or dialogue, the figures were only capable of dancing, and further, were fixed to a box upon which they danced

* Byrom concedes, from his own investigations, that these figures were internally strung, but believes they were operated from above and moved about the “stage”

* Purschke believes that they were internally strung and operated by pulling the strings horizontally from inside the box

* Byrom believes that the Roman tradition of puppet theatre continued underground by wandering entertainers after the fall of Rome

* Purschke believes that after the fall of Rome, puppet theatre (if there were such in those times) died out completely, as there is no mention of  real marionettes until the 16th century

An easy way of determining something definite is to avoid the error of translating Greek and Latin references to string-pulled figures as “puppets” or “marionettes” and their operators as “puppeteers”. For instance, the Greek neurospasta and the Roman sigillario are normally translated into English as “puppets” or “marionettes,” and neurospaston as “puppeteer”. This procedure causes no end of problems in attempting to prove or disprove the contention that the Greeks and Romans possessed a puppet theatre as we know it today. Byrom takes advantage of this in attempting to prove the ancients had a real puppet theatre. He simply follows the old way and translates neurospasta and sigillario as “puppets” or “marionettes,” and of course one is easily lead from this to his point of view. But to be fair, he is not alone in this. Practically every translator, whether a puppet historian or not, translated these passages using our modern words “puppet” or “marionette”. This practice seems to go back as far as the Renaissance, when the ancient texts were rediscovered by the Italian humanists. This only shows us that marionettes were known at the time of the translations. But the literal translation of the Greek neurospasta (and like words) is “string-pulled,” and the Latin word sigilla is, in any Latin-English dictionary, defined as a small statue or figurine only. However, the contexts of some of the passages definitely indicate motivated figures of some kind. With this in mind, we will attempt to determine whether or not there was indeed a puppet theatre as we know such today in ancient Greece and Rome.


An important text for this whole controversy is that of pseudo-Aristotle called De Mundo. This text is now known to have been written by a 2nd century author named Lucius Apuleius. (Helen Hagan, Moroccan Anthropologist, says "Apuleius was not a Roman, but a North African educated in Latin and Greek. He is the very famous author of The Golden Ass, among other works."   For further fascinating information, refer to her website concerning Apuleius, at: Many thanks to Ms. Hagan for the information.) De Mundo was a translation into Latin of an original Greek work, Peri Kosmon, once attributed to Aristotle. This reference has been given an honored place by historians of the puppet theatre. The Latin translation follows:

Those who direct the movements of the little wooden human figures [ligneolis hominum], when they pull the string controlling a limb, the neck bends, the head nods, the eyes look, and hands lend themselves to any action required and the whole body moves gracefully like a living thing.

This quote is a paraphrasing of the original text. The original Greek is translated into English as follows: “...just as the cord pullers [cordas trahunt], by pulling a single string make the neck and hand and shoulder and eye and sometimes all the parts of the figure move with a certain harmony...” George Speaight says that “this work was freely translated,” and “the original Greek, which was probably written in about A.D. 100, refers not to puppet showmen, but to ‘machinists’, and the whole passage probably alludes to automata rather than ordinary puppets.”

The history and attribution of De Mundo is confusing, to say the least. There is even an entire book on all the textual history of this work entitled The Text Tradition of Pseudo-Aristotle “De Mundo,” by W. L. Lormier. We can safely say that Apuleius was translating a passage from Greek to Latin in terms of the “string-pullings” of his day, but the original Greek does not necessarily refer to real marionettes. The original Greek text mentions only one string, while his translation (commentary?) implies a few strings, one for each part that was moved. Today’s puppet historians use this text as proof of marionettes in Roman times, but they are looking at the words through modern eyes. The strings must have been threaded inside the figure itself (as shown above), and not strung from above as with today’s marionettes. The texts, if read without modern preconceptions, do not necessarily invoke images of true marionettes.

An interesting comment in this regard (in French) is by Bernard Suzanne, at , where he says, about Plato’s Allegory of the Cave (many thanks to Mr. Suzanne for his invaluable assistance with the translation):

The French “Les faiseurs de prodiges” [English: “the wonder makers”] translates as the Greek tois thaumatopoiois, a compound word in which can be found the word thauma (plural thaumata), “prodige, merveille, objet d'étonnement” [in French; “wonder, marvel” in English]. This word is undoubtedly not chosen by chance, since in the Theaetetus, 155d2-4, Socrates makes thaumazein, the act of wondering, the origin of philosophy. The “phenomena” that we see must provoke our wonderment if we are to start looking behind appearances to one day have a chance of beginning our journey toward light. That is why translating the Greek thaumatopoiois in French as “montreurs de marionnettes” [English: puppeteers], as do most translators ([in French:] E. Chambry, Budé ; L. Robin, Pléiade ; R. Baccou, Garnier ; B. Piettre, Nathan ; M. Dixsaut, Bordas ; P. Pachet, Folio ; T. Karsenti/Y. Prélorentzos, Hatier ; or, in English, “exhibitors of puppet-shows,” P. Shorey, Loeb ; “puppeteers,” G. M. A. Grube, Hackett ; “puppet-handlers,” A. Bloom, Basic Books), using the context to select from all the possible meanings of the Greek word a specialized meaning and rendering it in French, or English, by a word which doesn't carry the other meanings of the Greek word as well, is to “trivialize” a text which leaves nothing to chance and deprives the reader of “resonances” precisely meant to arouse his wonderment.


The Greek “hôsper tois thaumatopoiois pro tôn anthrôpôn prokeitai ta paraphragmata, huper hôn ta thaumata deiknuasin,” I translate [in French] as “semblables aux palissades placées devant les hommes par les faiseurs de prodiges, par dessus lesquels ils font voir leurs prodiges” [English: “similar to those fences put in front of men by wonder makers, above which they show their wonders”.]

None of the ancient Greek and Roman texts can be used to demonstrate conclusively that a real puppet theatre existed in those times. Even the passage by Athenaeus (see my Chapter One), mentioning Potheinos the string-puller performing in the theatre of Dionysius, cannot prove the existence of actual puppet theatre, although this passage is often used just for that purpose. About this passage, Byrom says the following: “A puppet theatre as we know it there must have been; indeed we have the well known testimony of the physician Athenaeus in confirmation of this....” But Potheinos could have been displaying automata, as Heron of Alexandria and Philo of Byzantium (another mechanician like Heron) had done before him. Granted these shows were demonstrations before an audience by persons classed with entertainers, but there are no references to any characters acting as in a play - none are ever mentioned in any connection to motivated figures. They could have been rather like the department store window displays of animated figures that we see during Christmastime today. And, a single operator is almost always mentioned, so a simple and easily portable entertainment must have been the norm. In W. Beare’s The Roman Stage, 1964, quoting from Heron of Alexandria (c. 100 AD), we find wonderful and clear images of the “wonder show” of antiquity. Beare says “This seems to have been a box mounted on a pillar. The box had folding doors, which, when opened, revealed moveable figures seen against a painted background.” Heron’s words follow.

The problem is to make the puppet show [original Greek word(s) not available] open of its own accord and reveal the figures inside in motion befitting the theme of the story; then the pinax [panel] is to close automatically and after a short interval to open again, revealing other figures, some or all of which are if possible to be in movement. This process is to be repeated several times. The arrangement employed by the early designers was simple. When the pinax was opened, there appeared in it a painted head. This moved its eyes, raising and lowering them repeatedly. The pinax was shut and then opened again; the head had disappeared, but painted figures were seen arranged in accordance with some story. The pinax would shut and open once more, revealing another arrangement of figures to complete the tale. There were thus only three different movements: that of the doors, that of the eyes and that of the curtains.

But in our time designers have introduced interesting tales into their puppet-shows [original Greek word(s) not available], and have made use of many and varied movements. As I promised, I will describe one show which seemed to me to be the best. The story set forth in it was that of Nauplios. This is how it was divided up:

At the beginning the pinax opened and displayed twelve painted figures, arranged in three rows. These represented some of  the Greeks preparing their ships and getting ready to launch them. These figures moved, some sawing, some chopping…. They made a loud noise, as in real life. After a sufficient time had elapsed, the doors shut and opened again, and there was a new arrangement. It showed the ships being launched by the Greeks. The doors again shut and opened, and nothing was visible in the pinax except sky and sea. Presently the ships were seen sailing by…. Again the pinax shut and opened. There were now no ships to be seen, but Nauplios was there brandishing his torch and Athene standing beside him. A flame burned above the pinax…. Again it shut and opened, revealing the wreck of the fleet and Ajax swimming in the sea. A mechanism in the top of the pinax was raised, there was a peal of thunder, a lightning flash fell on Ajax, and his figure vanished. The pinax closed, and the story was ended.

Aristotle in his De Motus Animalum (On the Movements of Animals), Book 7, gives us some finer details of how the figures were moved in these shows.

The movements of animals may be compared with those of automata, which are set going [i.e., set in motion] on the occasion of a tiny movement; the levers are released, and strike the twisted strings against one another; or with the toy wagon. For the child mounts on it and moves it straight forward, and then again it is moved in a circle owing to its wheels being of unequal diameter (the smaller acts like a centre on the same principle as the cylinders). Animals have parts of a similar kind, their organs, the sinewy tendons to wit and the bones; the bones are like the wooden levers in the automaton, and the iron; the tendons are like the strings, for when these are tightened or leased movement begins.

Allusions to the concealment of the demonstrator are also to be found, but these can also be interpreted to refer to the non-interference of the showmen, who never seem to touch the figures directly to make them move, and possibly also to the concealment of the figures’ means of motion.

For as in the [wonder show, thaumati] all those things which are visible are inanimate while that which works the [string-pullings, neurospastei] is invisible …. In the same way, the Creator of the world sends his powers from an eternal and invisible place so that we are moved like [wonders, thauma] toward that which pertains to us, namely seed and procreation. (Philo, Quaestions et solutiones in Genesin, Bk 3, Sec 48.)

From him (God) they come, if indeed from him and not rather from Achamoth herself, from her in secret, he perceives nothing of her, and like a [figurine, sigillario] guided from the outside, he is moved in all his activities. (Tertullian, Adversus Valentinianos, Chap 18.)

Another passage as to the operation of certain moving statuettes is by the 4th century AD Roman Bishop of Ptolomais, Synesius, in his De Providentia: “And just as the [conjuror’s figures, praestigiatorum sigilla] which are moved by strings are still animated even when he who has given the impulse has stopped doing so, but are not animated for long because the source of motion is not incorporated in them so that they move only so long as the energy imparted to them persists or until it is dissipated by the continuance of the movement....” This likely describes a kind of automaton, which by the yank of a string, sets the figure in motion. This description of Synesius also calls to mind the passages above from Peri Kosmon, and Aristotle’s De Motus Animalum.

It is my belief that what we have seen here described by Heron and Aristotle, are exactly the neurospasta, sigilla, thaumata, etc., mentioned by the other writers of ancient Greece and Rome. More than this, a seamless continuity can then be demonstrated from ancient Greece and Rome to Medieval times for automated figures, as indeed this type (the cabinet of moving figures) were the earliest known in Medieval churches. It is probable that there were in existence, at the same time, less sophisticated neurospasta, performed by the wandering street entertainers. These, instead of using weights and counterbalances, likely utilized the simple pulling of strings to affect all the automated movements and transformations.


Alas, we are not here in the realm of certainty, so we must leave these investigations and hope that sometime in the future new evidence, perhaps actual ancient string-pulled figures, or new texts will be discovered that throw some much-needed light on this field of research.


Besides the word koree mentioned in my Chapter Three as being evidence of the possible existence of glove puppets in ancient Greece, the reference in Plato’s Allegory of the Cave to the figures over a screen by entertainers can also be used as further indication of this possibility (it is admitted the evidence is very slim). We can go further in that even the swazzle (in Italian pivetta) may have existed in ancient times (the swazzle is an English word denoting an instrument placed in the mouth of a glove puppeteer to alter his/her voice to a squeak, as for the well-known Punch character). The recent and valuable book by Henryk Jurkowski, A History of European Puppetry from its Origins to the End of the 19th Century, tells us that such devices were not known before the beginning of the seventeenth century, according to the research of Frank Proschan. However, Duchartre, in his famous book on the Commedia dell’Arte, The Italian Comedy, states that “It was from Maccus also that Pulcinella (one of the Neapolitan Pulicinellas) acquired the habit of peeping like a frightened chick - a sound which his father [meaning Maccus] had exaggerated by means of a sgherlo or pivetta.” Maccus, a character of the Attellan farces of ancient Rome, is considered by many historians of the theatre to be the ancient Roman ancestor of Punch, or Pulcinella, as he was originally named.

Of a more definite nature, Victor H. Mair in Painting and Performance, page 64, says “In China, from at least the eleventh century, performers who worked articulated puppets made them speak in a shrill, nasal voice.” In a footnote to this on page 211, Mair mentions “In the T’ang-yin pi-shih there is a particular legal case that cites Shen Kua (1030-1094), Meng-hsi pi-t’an, fascicle 13, for this information about the whining, nasal quality of puppets.” In addition, we have the following from the researches of John Cohen in his Human Robots in Myth and Science:

...they [the teraphim, household idols mentioned in the Old Testament] may have resembled the ob and yidde’oni, used by Biblical necromancers, and considered by some to be phantoms under celestial influence which one could converse with men and give them counsel. Ob is one who speaks from between the joints of his body and his elbow joints, while yidde’oni is one who places the bone of a yid’oa’ (a beast or bird) in his mouth where it speaks of itself.

The ob seems then to be some sort of ventriloquism, but yid’oa’ sounds suspiciously like the swazzle, especially as there is mention of a bird, for which use the swazzle is sometimes put, i.e., to imitate bird whistles.

If the above confirms the existence of a swazzle instrument in ancient times, then we would have here practically a proof that glove puppets existed in those times, as the swazzle was associated almost exclusively with glove puppeteers, and, in addition, that dialogue between puppet characters was used in ancient Rome, and, that therefore there existed acting and plays in a real puppet theatre. But we are here in the realm of pure speculation. There are hints rather than proofs. Suffice it to say that the swazzle instrument was in use by the Chinese since at least the 11th century.


That puppets were known by prehistoric man is proved by the fact that articulated figurines, many operated by the pulling of strings, were found already existing in the New World (North and South America) upon the arrival of the Europeans. Further, actual artifactual proof of the existence of articulated figures in prehistory can be found in the 26,000 year-old ivory figurine, now possessing only one articulated arm, that is actually nicknamed “the marionette” by archaeologists. This figurine was discovered in Europe, in Brno, Czech Republic. (See  “The Mammoth Ivory Male statuette from Brno” for details and pictures, and also The National Geographic Magazine, Vol 174, No 4, October 1988.) 

Bil Baird in his book The Art of the Puppet tells of Amerindian masks with jaws that open and shut by pulling strings. The fact that nearly all peoples worldwide had puppets and masks (used in ceremonies and rituals) since antiquity, indicates that prehistoric man in his ancient migrations around the globe brought these arts along with him. So puppetry is likely over 100,000 years old!

The first puppets were ritual figures. There is not much evidence to show how moveable idols were used in ancient times, but a quote from the Old Testament may assist us in visualizing this. In Habakkuk 2:19 (7th century BC):

Woe unto him that saith to the wood, Awake; to the dumb stone, Arise, it shall teach! Behold, it is laid over with gold and silver, and there is no breath at all in the midst of it.

A Bible commentary by John Calvin (1509 – 1564) gives a more literal translation of this passage, followed by an explanation by the editor:

Woe to him who saith to the wood, "Awake, Arise;”
To the dumb stone, "It will teach:”
Behold, it is covered with gold and silver!
Yet there is no breath within it.

The two verbs, "Awake, Arise,” stand connected with "wood,” [in Calvin’s original Latin source] and they are so given in the Septuagint; and there is a striking contrast between the dumb stone and teaching. -- Ed.

So the ancient priests in the temples of the Middle East spoke to their idols made of wood, imploring them to awake and move. Which they likely did, as moveable idols and statues were known since far back before this time. In fact, in the Encyclopaedia of Religion and Ethics (1908-1926), by James Hastings, editor, Vol. IV, page 792 – 796, under “Divination (Egyptian),” we find much information about the ancient Egyptian moveable idols.

Recent Egyptological discoveries show that all traits of this curious ceremonial [of Ethiopia, previously described by the writer of the article, George Foucart] were borrowed by Ethiopia from the divinatory ritual of Egypt. On the tomb of Nib Uonnaf at Gurneh (see Sethe, Z.A. xliii [1906] 30 ff.) there is an account of the election of the high priest of Amon-Ra. The candidates were led before the statue of the god. They were all shown to it in turn, ‘and not to a single one of them did it make the motion hanu [see below for the actual gesture] except, said the King, when I pronounce thy name.’ Then, Nib Uonnaf being thus chosen, the statue conferred the power upon him by four magic passes. A second text, discovered later, proves that the custom was in existence even in the time of Amenhotep III., and it is quite logical to suppose that it goes back to a much earlier period; it may perhaps be even as ancient as the worship of the god himself.

Foucart ponders on the meaning of the motion hanu, that it “may have been a movement of the arm of a jointed statue, accompanied perhaps by a sound, a whistling, or a cry, of suitable strength,”  and that “even though hanu means a shaking of the head, the statue certainly moved or stretched out its arm….” He goes on to say that

[G.] Maspero, in all the works in which he discusses these ‘prophetic statues’ (…), holds that they were actual jointed dolls, with strings attached to their arms and heads, and that the officiating priest pulled a string for each response and each gesture. In his earliest works (…) he even seems to admit the existence of actual machinery, worked, when required, by fire or steam. The explanation that the statue had a jointed head seems to be generally accepted.

In Japan, even today, we find puppets still being used in rituals. A survey of these was published by Jane Marie Law in her Puppets of Nostalgia, 1997, which may assist us in imagining the puppet ritual ceremonies of early civilizations.

* a substitute to remove pollution from persons (Hitogata)

* substitutes for fetuses, infants, and children to protect them from evil influences and disease (Amagatsu and Hoko [or Boko])

* as an effigy or ritual substitute for a dead child (Kokeshi)

* in rites for aborted or miscarried children (Mizuko Jizo)

* washing the body of the bodhisattva (a “being of enlightenment,” essentially a “Buddha to be”) in effigy as an act of pious devotion. “The merit of this action can be transferred to the practitioner in the form of a healing” (Law).

* naked puppets as imitative magic in fertility rites (in Noroma performances of ningyo [dolls, puppets])

There was no puppet theatre (plays with dialogue and/or narration) until perhaps Medieval times. The earliest mention of string-pulled figurines is by Herodotus in Greece c. 400 BC (see my Chapter One) and this was in the context of an Egyptian procession, not a puppet play. Many figurines were found during excavations of ancient civilizations which possessed articulated limbs. These have been investigated and it seems the earliest were used in rituals and ceremonies, either of a religious or magical nature. Another possibility was that they were amulets and charms - or perhaps all of these together. It is even possible that the articulated figures of ancient Greece and Rome with stiff wires extending out of the crowns of their heads were hung from the necks of men and women as amulets. One can imagine the hooked ends of the rods attached to a chain or string to be placed around the neck as a good-luck charm. The ancient Greek and Roman figurines were certainly small enough for such use. (Of course, the relation between these and the modern-day Sicilian opera dei pupi, which still use an iron rod at the heads to move the puppets, cannot be denied, but a direct line of descent is not discernable between these figures in antiquity and the puppets of the Renaissance. The opera die pupi style of show is generally thought to have originated in Italy during the 19th century, or slightly before.) Small dolls have been used as charms since ancient times by shamans, warriors, and soldiers and were tied to their persons to offer protection from evil influences and for good luck during their travels and forays into other lands.

Later, of course, dolls developed as children’s playthings. We see many of these in ancient Egypt, India, and then Greece. Along with the human theatre developed the puppet theatre, starting in ancient Greece and the Far East. However, in the Far East, including India, human theatre developed along quite different lines in comparison to the West. At the start of the Christian era, human theatre in the Orient was still at a pantomimic stage compared to Europe, i.e., mimed acting with narration provided. Many folk theatre forms in Asia even today are of this type.

Much can be gleaned about the sources of puppets and dolls in different countries by looking at the etymology of the words used to refer to them. The first and most obvious would be the Medieval English word “poppet,” which originally meant  “doll” (c. 1413), and was a reflex of the French poupette, of the same meaning. According to David A. Pharies in his article “The Etymology of the Spanish Titere ‘Puppet’,” published Fall 1985 in the Journal of Hispanic Philology, Vol X, no.1,

The later form puppet was for a time ambiguous (‘puppet’ 1538, ‘doll’ 1562), […], but by the end of the 19th century has lost the etymological meaning.

Besides Pharies’ article showing that the English “puppet” (actually a doll) was derived from France, he explains that the Spanish titere for “puppet” was derived from the French petite, “little one”. The oldest reference to titere in Spanish appeared in the writings of Bernal Diaz del Castillo, ca. 1560. The English word “doll” came into general use only in the 18th century, and was likely derived from the nickname for “Dorothy,” which was “Doll,” or the German “Docke” for a doll. This last could be a sign of heavy German influence on English dolls, causing the name for them to change from “puppets” to “dolls”.


As the French poupette was derived from the Latin pupa, meaning “doll,” it seems the path of diffusion of dolls was from ancient Rome to France (and Spain?), then to England. Ultimately, the use of dolls was a legacy of the Roman occupations of Europe. Although puppets were known in Europe since at least the 12th century, there was no single word for stringed puppets until perhaps the 17th century, when the word “marionette,” meaning “little Marie,” was applied to these theatrical figures. There is some disagreement whether the word was originally French marionnette or Italian marionetta, however, the word in France means all types of puppetry, whereas in Italy and England only the stringed variety. This possibly indicates that the English and Italians had seen French puppeteers arrive in their countries with stringed puppets, so borrowed the word to designate that type only. This kind of thinking can be applied to the first English mentions of the Italian Commedia dell’Arte character Pulcinella (later Anglicized to “Punchinello” then shortened to “Punch” in England), in Pepys’ Diary of the 17th century spelled variously “Polichinelli” or “Polichinello,” referring to puppet shows he had seen featuring this personage. The spelling is very close to the French “Polichinelle,” indicating that the Italian puppet character came to England from Italy via France.

In The Lost Civilizations of the Stone Age, by Richard Rudgley, 2000, there are outlined two words of great interest that, according to some linguists, existed in the proto-language, the Mother Tongue as it were, of prehistory, used perhaps 15,000 years ago. These are KUNA for “woman,” and PUTI for “vulva”. Many languages of the world today show these roots in words of the same or similar meaning. KUNA also appears in some form in various words for puppet and/or doll, such as in kukla (Turkish, Russian, Greek), ku’ei lei (Chinese), and kundhei (Oriya in India). This may indicate that the first dolls and puppets were originally female figures, perhaps going back to the almost universal prehistoric Mother Goddess cult. PUTI appears in various languages as puppet (English), poupée (French), Puppe (German), pupa (Italian), putrika, puttali, puttalika, putul, putli (various languages in India), etc. The meaning “vulva” in the root PUTI is perhaps another throwback to the Mother Goddesses of prehistory. As indicated by the reference to a woman’s sexual organ, it is even possible that before history, women, perhaps even temple dancers and courtesans, were engaged in puppet playing, or at least, were connected to puppet performances in some way. This would explain some of the contempt, even hatred, shown puppets and puppet shows in most of the world today. It could go back to the old contempt and hatred for womankind and the practitioners of the “oldest profession.” In ancient times courtesans and temple dancers were also entertainers. In the book Bunraku, Japan’s Unique Puppet Theatre, 1964, Shuzaburo Hironaga says in this regard

At the dawn of civilization in Japan, maidens in the service of Shinto shrines operated a pair of simple dolls, male and female, as they recited a prayer. Around the 7th century wandering puppeteers migrated to Japan from the Asian continent via Korea. Treated as hinin (outcasts), they toured from one province to another in the country and were known by the name of kairaishi or kugutsumawashi.

These travelling entertainers carried in front of them a wooden box hanging by a strap from the neck. Visiting houses from door to door, they took finger type dolls out of the box and made them dance as a means of asking for alms. Female kugutsumawashi often engaged in prostitution on the side.

One should not forget too that some shamans practicing in primitive societies are women, and many shamans use doll figures in their rites. Another interesting fact is that Pulcinella (the original Italian name of Punch) actually possesses the feminine suffix.


Other word correspondences are bombe (or gombe) in India, compared to bambola in Italian; and pava in Kerala, India, compared to bavastel in medieval Spanish. Then we have bastaxi from medieval French, the Chinese poh-tai-hsi (or -xi), and the Taiwanese budaixi. One should remember what many scholars have already pointed out: namely, that such word correspondences can be continued ad infinitum, so we should not rely too much on them to prove our theories. Their best use is when we can demonstrate the correspondences are in sound and meaning both together, as is the case in these examples.

A fascinating object for study, related to puppetry, is picture storytelling. This form of entertainment was very prevalent in Europe during the Renaissance and after, especially in Germany, where such a showman was called Bankelsanger (bench-singer). Generally, these entertainers stood on a wooden bench (in order to be seen in a large gathering) while telling stories in front of a pictorial sheet, pointing with a stick to the illustrations, sometimes with an assistant who sings or who sells the stories in printed form, perhaps also playing a musical instrument. Often, the storyteller took on some of these activities and the assistant helped in other ways, such as playing the music, if any. In Italy, street entertainers were called cantambanco, also meaning “bench-singer,” and, we have entertainers known as mountebanks, meaning “mount-on-bench,” both obviously related to the German Bankelsangers, or bench-singers.

The earliest illustration of such an entertainment is in fact German. See Victor H. Mair’s Painting and Performance, figure 28, p. 151 for the engraving dated c. 1480 – 1490. But there survives in museums and private collections remains of even earlier forms of this type of storytelling.  Mair mentions the exultet rolls from South Italy, dated from the tenth to the twelfth centuries. These were vertically unrolled from the church pulpit during Easter vigil services. Another possible early evidence of such picture storytelling, never heretofore mentioned by researchers in this regard, is the Bayeaux Tapestry in England, showing the Battle of Hastings in 1066. See the website for details and history of this artifact. This tapestry has all the features of the Indian, Chinese, and Indonesian forms of picture storytelling, but it is an embroidered cloth, and extremely long. Another parallel to what Mair has brought out about these entertainments, is the fact that the Bayeaux Tapestry also has trees separating the various scenes, just as in the Javanese wayang beber and the sole surviving Chinese T’ang Dynasty picture storytelling scroll. I am of the opinion that this cloth was taken out on ceremonious occasions and the story of the Battle of Hastings was recited as the cloth was unrolled. Such a very long cloth (it is approximately 70 metres long and half a metre wide) could not have been displayed conveniently otherwise. After this period, there seems to be no indication of the existence of picture narration until the 15th century in Germany, when it became widespread all over Europe. How do we account for this?

Mair has proved that picture storytelling itself originated in India in the centuries before Christ, and diffused eastward to China, southeast to Indonesia (wayang beber), and westward to Europe. It seems then that migrations of Indians were the method of transmission, but Mair does not give any details of how this occurred in the case of Europe. For instance, if Indians brought picture narration to Europe, how is it that Europeans took it up, and even the Catholic Church?  The very early appearance of this form in Southern Italy as a church practice may be the result of its transmission from the Byzantine Empire (in Turkey), once the principle seat of the Catholic Church. Gypsy tribes from India were known to have arrived in Byzantium c.1000 AD, which is in line with the dating of the exultet rolls. It is also known that Gypsies were entertainers, but it is not known if among them were picture narrators, though this is likely. The Gypsy tribes came from Northern India, and this is the place in India where we find the earliest manifestations and longest still-existing traditions. But this does not solve the riddle of how the Catholic Church took it up from these foreigners who at first did not speak the local languages.

The forms of picture storytelling transmitted from India to China and Indonesia as outlined by Mair show that the Chinese and Indonesian forms were of one type, and, in fact, the wayang beber of Java and Bali is very similar in form to the only surviving such Chinese scroll (dating to the 9th century). From history we know it was clearly not the Gypsy tribes that were responsible for transmitting picture storytelling to these two countries. The mode of transmission to these places was almost certainly the propagation of Buddhism from India.

So then, the only other method of transmission of picture narration from India to Europe must have been via the Gypsies. To explain how the Europeans took it up, it is probable that when the Gypsies arrived in Byzantium, they had to earn a living with their entertainments, so they quickly had to learn the local languages to do so. They most probably entertained the local poor folk like themselves, so Byzantine street entertainers eventually learned the technique from the local Gypsies. Later, after further migrations to Europe, the Gypsies were looked down upon, hated, persecuted, and forcibly driven out from Europe (not entirely successfully), and what remained were the native, European, entertainers’ forms of picture narration. In fact, many of the engravings of European picture narrators in Mair’s book show peoples who look decidedly like Gypsies. Picture storytelling must have become quite common for the church to have then taken it up in the form of the exultet rolls. When the seat of the Catholic Church moved from Byzantium to Rome, likely went the practice of the exultet rolls. This later developed into carved or painted wooden boards used for religious storytelling that were originally placed behind the church altars: the well-known retablo in Spain, retable in France, and reredos in England. (As a note of interest, “reredos” must have been the original name of the “raree show,” an obvious mispronunciation, and actually the same type of entertainment.) I suspect this is where picture storytelling in the Catholic churches “disappeared” to after the last use of  exultet rolls in 12th century Italian Easter vigils. Anne Pellowski, in her book The World of Storytelling, 1977, page 71, speculates similarly. She says

A case might be made for the possible transformation of the exultet roll into the early form of triptychs or trunk altars that may or may not have been the forerunners of religious picture sheets used by early bankelsanger.

As with puppetry, the poor, illiterate street entertainers doubtless continued displaying their picture sheets to the public, being of such a class as to be unworthy of notice by the chroniclers of the time. It is known that these church entertainments later were frowned upon by church officials and so were eventually thrown into the streets. The retablo church furnishings then passed into the hands of the street entertainers. The stationary figures of the carved retablos before that time had became moveable, even when they were used formerly in demonstrations in the churches of Europe.


An early indication of the possible existence of marionettes is by William Lambarde of England, a 16th century antiquary, describing a play that included a puppet character called “Jack Snacker of Whitney”. Speaight thinks it “was almost certainly being performed by 1500” (p. 53).

In the dayes of ceremonial religion they used at Wytney to set foorthe yearly in maner of a Shew, or Enterlude, the Resurrection of our Lord and Savior Chryste, partly of Purpose to draw thyther some Concourse of People that might spend their Money in the Towne, but cheiflie to allure by pleasant Spectacle the common Sort to the Liking of Popishe Maumetrie; for the which Purpose, and the more lyvely thearby to exhibite to the Eye the hole Action of the Resurrection, the Priestes garnished out certein smalle Puppets, representinge the Parson of the Christe, the watchmen, Marie, and others, amongest the which one bare the Parte of a wakinge Watcheman, who (espiinge Christ to arise) made continual Noyce, like to the Sound that is caused by the Metinge of two styckes, and was therof commonly called, Jack Snacker of Wytney” [see Scott Cutler Shershow, Puppets and ‘Popular’ Culture, 1995].

My interpretation of this text is that the jaws of the puppet Jack clacked together in fright at seeing Christ rise from the dead. But this puppet could have been operated by any number of means – no strings are mentioned.

An even earlier evidence of puppet plays is also from England. George Speaight has discovered an article in the Records of Early English Drama, issue 1979:2, wherein Ian Lancashire has connected a 1431 English puppet play Joly Walte and Malkyng to the oldest English play-text Interludium de Clerico at Puella (The Interlude of the Cleric and the Girl), dated c. 1300 AD. Speaight goes further and thinks it a possibility that the 1300 fragment is also a puppet play text! (See his The Earliest English Puppet Play?, 1997, published by Ray DaSilva Puppet Books.) His reasons for thinking this are that the play (in its surviving fragmentary form), has only two characters appearing on the stage at a time - this is a necessary convention of  traveling glove-puppet shows, where the single puppeteer slips a puppet on each hand, allowing only two characters on stage for each scene – plus, the fact that “Malkin” (the name of the “Girl” character in this play) had, among other meanings, also the sense of “puppet,” according to old English dictionaries.

An entry in the Annals of English Drama, 975 – 1700, 1964 edition, lists a liturgical “Play or Puppet Show” dated 1491 of the well-known Quem Quaeritis, where the three Marys come to Jesus’ tomb on the third day following His crucifixion only to find the stone before the tomb entrance rolled away and an angel standing in the doorway. The angel asks them

Whom seek ye in the sepulchre, O Christians?

Jesus of Nazareth, who was crucified, O angel.

He is not here, He has arisen as He foretold:

Go, announce that He has arisen from the grave.

This type of play (called a Trope) is known from a 10th-century manuscript from the monastery of St. Gall in Switzerland, the author most likely being the Benedictine monk Tutillo. Whether the Quem Quaeritis was performed with puppets as early as the 10th century is unknown. In France, in 1408, a certain Perrinet Sanson performed with puppets in a French village, according to George Speaight, whose source was Domino Du Cange’s Latin glossary of 1883, Glossarium Mediae et Infimae Latinitatis under “bastaxius.” There are other references to puppeteers listed by Du Cange for around the same century. Finally, I have already mentioned in my Chapter Four the Provencal romance Flamenca of the early 13th century.

As to the question of whether or not dialogue between the puppets was used in the medieval puppet shows, we refer the reader to the two famous Li Romans du bon roi d’Alixandre illustrations of the 14th century (see my Chapter Four for these pictures). We note that no interpreter (the speaker for the puppets outside the booth) is visible. If he had existed in glove puppet shows in the 1300’s, then surely he would have been included in these illustrations; the illustrator would have seen him as an integral part of such shows. But he is nowhere to be found. This leads us inevitably to the conclusion that medieval hand puppets spoke for themselves through the hidden puppeteer. By referring to the particular Alixandre illustration that depicts the man and woman puppets having a conversation while three girls in the audience watch, we may also conclude that dialogue was a major part of the shows - with some slapstick added à la Punch and Judy, as we can also see the male puppet in a typical court jester’s costume with his cudgel resting on his shoulder, just as glove puppets have done since their birth long ages ago.


In my Chapter Two, I discuss the Emperor Wu (or Wu-ti) episode, and the fact that this particular legend is often cited to prove the existence of shadow theatre in China in 121 BC. I have lately come across a very recent complete study and re-evaluation of this legend. This is “Documentation Relating to the Origins of Chinese Shadow Puppet Theater,” by Alvin P. Cohen, in Asia Major, vol. XIII, part I, 2000. Prior to reading this article, I was ready to accept this legend from the Han Dynasty as based on fact after I learned that the writer of the story, Ssu-ma Ch’ien (also spelled Sze-ma Ts’ien), was a contemporary of Emperor Wu, and actually lived at his court. Regardless of whether or not Ssu-ma Ch’ien was contemporaneous with Emperor Wu, we require to know if his text is a reference to shadow theatre. Mr. Cohen points out that this story of Emperor Wu was repeated over and over again down the ages, with many variations, all without, except for one instance, specifically mentioning shadows. One would think that if shadow plays were known in China at any time prior to the Sung dynasty, real shadow plays would have been spoken of when recounting the Emperor Wu story.


The single mention of shadows in a Emperor Wu story comes from a version of the tale dated 983 AD, the Hsin-lun fragment in T’ai-pin’ing yu-lan, which was compiled by Li Fang and others. Mr. Cohen says that the phrase shen ying means either “spirit and shadow” or “spirit shadow”.


Li Shao-chun established (or placed) the spirit shadow (shen ying) of emperor Wu’s concubine Li inside a drapery (chang) and had the emperor gaze at a distance (wang) and see it.


As Mr. Cohen puts it, “None of the three versions of this fragment provides even a hint as to how the image was produced.” For myself, the late date of this version puts it fairly close to the same period as the first unambiguous references to Chinese shadow plays (11th or 12th century), so dispute on this point is unnecessary.

The early evidence for shadow theatre in India too is very ambiguous. I quote again Mr. Cohen’s article.


The shadow theaters of Java and India are well known, however the dates of the earliest references to this type of theatrical and its antecedents are shrouded in the uncertainty of the dating of sources as well as the interpretations of the apparently relevant text passages.

Scholars of Indian shadow theater often interpret certain passages in the epic Mahabharata and other early texts as referring to shadow theater, but the dating of the numerous layers within these texts and the interpretations of the relevant passages are debatable.

Other recent studies on Javanese, Arabian, Turkish, Chinese, and Indian shadow theatres have brought to light the fact that the earliest unambiguous references to shadow theatre in these lands are all of approximately the same period, around the 11th to the 13th century.

For Java, the poem Arjunawiwaha (Arjuna’s Wedding) is frequently brought forward to demonstrate that shadow theatre (wayang kulit) existed in Java in the 11th century (the word used in this poem is not wayang, but ringgit, which is a synonym for wayang, but can also mean a female dancer). The fact however that wayang can signify “phantom” or “ghost” (besides “shadow,” “play,” “drama,” “theatre,” or “puppet”) puts in doubt also the other, earlier texts using this word, or ringgit (see my Chapter Two for some of these texts). We must remember that the root meaning of various words denoting “mask” is exactly “phantom” or “ghost” (even in Europe), so the texts where wayang or ringgit are mentioned without any descriptions can also be placed in the category of references to masked dances (wayang topeng in Java).

A recent book on the subject of the wooden puppet theatre of Java, Voices of the Puppet Masters: The Wayang Golek Theater of Indonesia, by Mimi Herbert, 2002, has the following to say about roots of the word wayang.

The history of the wayang is elusive, but the name itself offers some important clues: the word wayang is most likely derived from yang, eyang, and hyang, referring to ancestors and sometimes deities, and from bayang, which means “shadow.” The veneration of ancestors and the propitiation of the gods are recurrent themes in the wayang tradition, and “shadow” could be a reference both to the early shadow puppets and to the spirits.

Even though the poem Arjunawiwaha is unambiguous in its reference to shadow puppetry, there are problems with the integrity of the text, as even P. J. Zoetmulder points out in his Kalangwan: A Survey of Old Javanese Literature, 1974, page 66. Here he also states that this poem as it has come down to us today is from the hand of a single person, in point of time before the end of the Majapahit period of Java (1292-c.1500).

The Javanese tradition of the origin of their shadow plays is that Sunan Kalijaga, one of the nine Muslim saints (Wali Songo) responsible for the propagation of Islam in Java, created or brought shadow plays to Java in the 14th century. Shadow play was apparently used to teach the principles of the new religion to the (mostly of Hindu religion) Indonesians. It must be mentioned that, in the main, Islam was passed to Indonesia by Muslims from south India and Gujarat, not directly from Arabia, although, in fact, the Arabs knew shadow plays before the 14th century, as will be discussed below.

What are the earliest unambiguous references to shadow theatre?


See my Chapter Two for the quoted text from Ceylon (Sri Lanka), given by Mair and also Ananda K. Coomaraswarmy (the source of Mair’s quote), as dating from the 12th century. For myself, “leather figures” is almost certainly referring to shadow puppets. However, Stuart Blackburn in his 1996 book, Inside the Drama House, states that the only pre-twentieth century evidence he has located for Indian shadow puppetry is “the inclusion of ‘leather puppet play’ [tol-pava-kuttu]  in a list of performing arts and amusements in a mid-eighteenth century poem by Kuncan Nambiyar.” He considers the Ceylonese and other texts unreliable because the literary evidence “is so sparse, often ambiguous, and lacks corroboration from inscriptions,” meaning they are difficult to date with confidence.

I have recently investigated the text from Ceylon, the Mahavamsa (see my Chapter Two for the text itself), using Wilh. Geiger’s article “The Trustworthiness of the Mahavamsa,” in The Indian Historical Quarterly, Vol. VI, No.2, 1930. This text is a Ceylonese chronicle, divided into two parts, the Mahavamsa (Great Chronicle), followed by the Culavamsa (Little Chronicle). Each treats a different period of history, in chronological order, and each was written by a different author or authors. It is the dating of the first part of the Culavamsa that we are concerned with here. Geiger dates the authorship of this section c.1211 AD.

In the Mahabharata, the reference to rupopajivana (which translates as “depictions of life through figures or characters”) cannot be dated or explained as written except for the commentary by Nilakhantha, which explains it in 17th century terms, proving only that shadow theatre existed in India in the 17th century. (One author in the Sangeet Natak puppetry issue number 98, 1990, gives the date of this commentary as the 12th century, and a second author in the same issue gives the 17th century. Most scholars use the 17th.) An alternate translation than that I have given in my Chapter Two follows. (The bracketed words, except for rupopajivana, are insertions by the article’s author, M. Naghabhushana Sarma.)


Presently [depictions of] life through characters [rupopajivana] known as jala mandapika is popular in South India. It shows leather characters behind a curtain [in tales] containing the deeds of kings and ministers. 


The intent of this commentary is now clearer. Nilakhantha (Neelakantha Panditha in Sarma’s article) seems to be saying that the type of puppets known in South India are shadow figures, as opposed to North India where they are (string?) puppets. The meaning of rupopajivana seems to be “puppets,” as rupa in Sanskrit means “form” or “figure.” I have discovered recently that this word rupopajivana appears in the 12th book of the Mahabharata, and Victor Mair in his Painting and Performance, gives an English translation of the entire passage with the meaning of rupopajivana as “puppets” (again see my Chapter Two for this). However, he does not show the original Sanskrit word rupopajivana, so I was for a long time not connecting rupopajivana with this passage in the Mahabharata!  The dating of the various sections of the Mahabharata is extremely difficult, if not impossible, so we are left with only Nilakhantha’s commentary. It seems the ancient Indians, unlike the Chinese, were not interested in historical dates or recording their chronological histories.

Recently, some Indian scholars have written articles (again see Sangeet Natak, number 98) noting that the earliest evidences of their particular state’s shadow theatre are from the 12th century. These references are from other, more obscure Indian texts (obscure to European scholars, in any case). Sarma mentions that this theatrical, because of its use in major religious festivals, must have been already a well-established entertainment by that time. So he infers that it must be somewhat older than the 12th century, although there is no other evidence to show exactly when it was created. Most scholars today give a wide range for the dating of these old texts.

The most reliable evidence we have is the existence of the Sanskrit drama Dutangada, which is expressly defined in the introduction to the play as a chaya-nataka, which literally translates as “shadow play”. Some scholars dispute this meaning. One instead believes it could mean “an epitomized adaptation of previous plays on the subject,” since “chayapajivin is one who composes poems which are reflections of other poet’s works” (S. K. De, “The Problem of the Mahanataka,” The Indian Historical Quarterly, Vol VII, No. 1, March, 1931). The Dutangada has been firmly dated by the scholars Cecil Bendall and Louis H. Gray as having been performed in Gujarat at a spring-festival held on March 7, 1243, in other words, in the 13th century. There are other later dramas also designated as chaya-natakas, and in one of these (Dharmabhyudaya) there is the following stage direction: “from the inner side of the curtain is to be placed a puppet (putraka) wearing the dress of an ascetic.” This makes it virtually certain that a chaya-nataka is really a shadow play.


We have already indicated above that the first certain mention of Chinese shadow plays (ying xi) is from the Sung Dynasty, in the 11th century (see my Chapter Two).  However, I have recently noted from Mr. Cohen’s researches (see above) that some of these texts are unreliable sources, so the earliest text that is authentic and reliable according to his studies is by Meng Yuan-lao (fl. 1110-1160). This then puts the earliest authentic reference to Chinese shadow theatre in the 12th century. (Mr. Cohen erroneous says “11th century” in his article.) The text is as follows (the bracketed words are by Mr. Cohen).

Since the Ch’ung-[ning] (1102-1106) and [Ta]-kuan (1107-1110) reign periods, among the entertainers and artisans of the pleasure-quarter of the capital, there have been…(the text gives seven names) shadow-theater [performers].

Southeast Asia

Early references, i.e., the inscriptions c. 840 and 907 AD, to wayang and ringgit are ambiguous (see my Chapter Two again) because in them there are no accompanying descriptions of these entertainments. Since these Javanese words can mean something other than “puppets” or “shadows,” we cannot use these inscriptions as indications of the existence of actual shadow theatre at that time. The first unambiguous reference is in a 11th century poem, the Arjunawiwaha, but as mentioned above, the text may have later additions, so we must date the wayang kulit somewhat after the 11th century, possibly the late 1300’s. The particular passage is:

Some, looking at the wayang [actually the word in the text is ringgit], cry, overcome with grief, though well aware it’s only carved leather engaged in acting.

The words “carved leather engaged in acting” here exactly describe a shadow play. But, leaving this text aside because of the difficulties caused by possible later additions which confound the dating of any particular section, we find the earliest mention of shadow theatre in Java is in the 14th century in the Malay Hikayat Raja-Raja Pasai (Chronicles of the Kings of Pasai), the oldest piece in the Malay language. It describes the coming of Islam in the 13th century to the now vanished kingdom of Pasai on the northern coast of Sumatera (Sumatra). For our purposes, there is also mention of various kinds of wayang.

The land of Majapahit was supporting a large population. Everywhere one went there were gongs and drums being beaten, people dancing to the strains of all kinds of loud music, entertainments of many kinds like the living theatre [wayang wong], the shadow-play [wayang kulit], masked plays [topeng], step-dancing and musical dramas. These were the commonest sights and went on day and night in the land of Majapahit. Food was in plentiful supply. Everywhere there were people going to and fro in numbers past counting. [See the Journal of the Malayan Branch of the Royal Asiatic Society (JMBRAS), vol. 33, pt.2, no. 190, June 1960, p. 103 in Malay and p. 161 in English, translated by A. H. Hill.]


A. H. Hill dates this section of the chronicle, part 3, to 1390 AD, and as the Majapahit period was pre-Islamic, we can conclude definitely that wayang kulit (leather shadow-puppet play) was known to the Javanese during the Hindu period. But none of this proves that the shadow play came from India, and does not exclude the possibility that the shadow play is indigenous to Southeast Asia, or that the Muslims passed this entertainment to the region. I mention this last possibility because evidence of Islamic graves in Sumatra at an early date show that Muslims were living in Southeast Asia as far back as the 13th century, and possibly before that.
The most certain evidence is from a group of bronze “zodiac beakers” from East Java, which show wayang-style figures around the outside. One in the Linden Museum in Stuttgart, Germany, is impressed with the Indian Shaka era date 1253, that is 1331 AD (14th century).

The Middle East: Arabia, Egypt, Persia, and Turkey

Shmuel Moreh, in his 1987 article “The Shadow Play (Khayal al-Zill) in the Light of Arabic Literature,” which can be found in the Journal of Arabic Literature, Volume XVIII, states that

The earliest Arab author to discuss systematically the technique of the shadow play, without however using the term khayal al-zill, was the scholar Ibn al-Haytham (d. 1039) [11th century] in his work on optics, Kitab al-Manazir. He defines khayal as “[translucent] figures [of characters and animals] which the mukhayyil <sic> moves so that their shadows appear upon the wall which is behind the curtain and upon the curtain itself.”

For Egypt, there exists three shadow play texts composed between 1260-77 (13th century) by Ibn Danyal, an Egyptian physician.

In the 14th century, Mongolian shadow players introduced the art to the Muslim Persians. Claire Holt in her Art in Indonesia, 1967,  page 131, says

Jacob [in “Das Chinesische Schattentheater,” no. 3 in Georg Jacob and Paul Kahle, eds., Das Orientalische Schattentheater, 1931] also mentions an interesting episode reported by the Persian, Dshuwani, connected with “wonderful Chinese plays, never seen before by anyone, behind a screen.” It was performed before the Mongol ruler Ogotai, third son and successor of Jenghiz Khan.

For Turkey, it is on record in the Tarih-i Misr [The Egyptian Chronicle] of Muhammed ibn Ahmet Ibn Iyas, an eyewitness to the events (according to the Turkish scholar Metin And), that shadow theatre was borrowed from Egypt in the 16th century. The context was a shadow play performed in the palace on Roda Island in the River Nile for the Sultan Selim, who afterwards told the performer, “When we go to Istanbul [Turkey], you will come with us so that my son too can see the shadow play.”


We are left from all of this with the information that shadow theatre was known throughout Asia since at least the 11th century. But it is uncertain where it was first known. The descriptions do not tell us if the practice was a new one, or an older tradition. To reach a reasonable conclusion on this question, we have applied some simple logic so that the probabilities of where it originated can then be located.

Today, many scholars assume that India is the originator of shadow plays. Even Mr. Cohen, in his article mentioned above, says

In spite of the problems associated with these early texts, there seems to be general agreement that the shadow theater originated in India and was transmitted to Java.

In a personal communication, the scholar Thomas Cooper of the University of California in Berkeley has pointed out that the distribution of transparent shadow figures and opaque ones in India seem to indicate that the transparent type seen in China must have migrated from the upper part of the Deccan in Southern India, where this type is concentrated, and which also happens to be the area of  Indian shadow plays that is closest to China. The opaque type is found in the extreme south(east) of India, and Indonesia, which country possesses the opaque type, is just below that point from India on the map. All this seems to indicate that the spread of both types is likely from south India going north to China and going south to Indonesia. Professor Cooper also observes that the Turkish and Arab types show a resemblance to the Chinese in that they are all of the translucent variety, and also they all utilize horizontally held control rods; the later spread of shadow theatre to the Middle East must have then been from China. Another bit of “proof” is that the historical migrations of shadow puppeteers within India itself is from the north (witness the Dutangada which was played in Gujarat in Northern India in the 13th century) to the south of that country, where we find the shadow theatre surviving today.

Mr. Cooper has also discovered that we can apply a well-known principal of historical linguistics to this problem of the place of origin of the shadow play. This principle is called “greatest diversity” and Merritt Ruhlen in The Origin of Language, 1994, page 163, says, speaking of an African homeland for the world’s languages, “one expects the greatest diversity in those areas that have been inhabited the longest, and in which, accordingly, variant forms will have had the longest time to emerge and accumulate.” In India we find the greatest diversity of shadow play forms. All the other areas have one basic type. For instance, China, Turkey, and Egypt have articulated single-figure translucent dyed leather types and not the opaque types, Indonesia possesses only opaque, articulated single figures (except in Cambodia and Thailand where articulated single-figure types exist alongside unarticulated composite “pictures”), whereas India has both the opaque and translucent types and many other variations besides, e.g., unarticulated figures (no moveable joints); round-shaped, transparent leather “pictures” (scenes); sometimes one performer handles all the puppets and speaks all the dialogues; sometimes many performers are present behind the screen; some have exclusively male puppeteers; others have both male and female performers; most use a portable screen, but one type, the tolpavakoothu of Kerala, uses permanent buildings to perform their plays; etc. This evidence of so much diversity strongly speaks for an Indian place of origin for the shadow play.

There is today general agreement among theatre researchers that the shadow play probably emerged from narrative picture scroll performances. Victor H. Mair has proved in his Painting and Performance that the earliest recorded occurrences of such scroll performances were in India, and this type of entertainment traveled around the world after spreading first to China and Indonesia. But it is difficult to form a clear picture how such an evolution from scroll pictures to the shadow play came about. After cutting out the figures from the scrolls, why place them behind a curtain? I am of the opinion that since shadow theatre was invented after puppetry in the late centuries BC, it may have been developed as an alternate form of puppetry. Plato in his Republic, mentions a screen behind which entertainers performed circa 400 BC (see my Chapter One). Today this is still the way hand puppets are presented, so it is not such a great step for puppeteers to remain behind their screen with their scroll cutouts. In addition, in India, the same entertainers often perform scroll presentations, shadow, and puppet plays.

Now we may again ask, when did the shadow play originate? The word saubhika, mentioned in my Chapter Two, is often brought out to prove the early existence of shadow plays in India. The first occurrence of this word is in the Mahabhasya, a grammatical treatise in Sanskrit by Patanjali, dated between 160 and 140 BC.  There is a highly controversial passage in this work, on which much ink has been spilled by scholars. The grammatical configuration of the word in this passage is sobhanika, and can be found at 3.1.26 in the treatise. It is a very complex and difficult text. There are many varied controversies that came out from this passage, but the only one we will consider here is the meaning of saubhikas, as a few scholars have translated it as “shadow players”. Many other scholars have also strongly disagreed with this idea. A good place to begin your own study of this passage is in Norvin Hein’s book The Miracle Plays of Mathura, 1972. He uses a text that is a very literal translation, plus any words of unagreed-upon meaning are left in Sanskrit. Of course, saubhika is one of them. This was in order to consider the matter afresh. His conclusion is that the saubhikas are entertainers, narrators, and dancers. I have approached the problem by looking at old Indian commentaries on this passage and definitions of the word saubhika. There are two conflicting definitions. The earlier one is by Haradatta in the 10th century AD who says that saubhikas mentioned in the Mahabhasya are teachers of actors.  The second definition of saubhika is by Somadevasuri also in the 10th century who defined it as “a displayer of forms of various sorts at night by means of a cloth screen.” This was a commentary on the word occurring in the Nitivakyamrita, in a list of professions suitable for employment as spies. Which definition should we follow? If we look at all known occurrences of the word saubhika and its variants in the ancient literature of India we discover that saubhikas were always classed with entertainers (magicians, musicians, acrobats, etc.). So the definition of Haradatta seems incorrect and we are left with Somadevasuri’s definition of them as shadow players. Considering that the first mention of saubhikas (sobhanika) was in the 2nd century BC (in the Mahabhasya mentioned above), it then follows that shadow plays were known in India by this time.


Most writers on early shadow plays in Europe mention the term ombres chinoises (Chinese shadows) that became current in France in the 18th century. But there are earlier records than the 18th century. My Chapter Two mentions a few. The recent book by Henryk Jurkowski, referred to above, has a few more. It has been mentioned often by scholars of the theatre that Italy first recorded this entertainment (for Europe), but I could not find the exact reference or source until Mr. Jurkowski’s book was published. In 1652, says Mr. Jurkowski (pages 94-95), the Jesuit Domenico Ottonelli of Italy published a treatise Della Christiana moderazione del teatro, where he says

Others are called bianti ombranti because they make their figures from pasteboard, and show them from behind a lighted cloth, thus from beyond only the shadows of these figures are seen, and this is the invention of Giuseppo Cavazza of Venice.

But the very earliest European mention of shadow figures was in 1619 in Spain. The name used was sombras chinescas, which translates exactly as “Chinese shadows,” the same expression as in France a century later. But there are indications of shadow plays in Europe even earlier than this. One is in a passage in The Suma Oriental of Tomé Pires (1512-1515), a Portuguese traveler, where he says

The land of Java is [a land] of mummers and masks of various kinds, and both men and women do thus. They have entertainments of dancing and stories; they mime; they wear mummers’ dresses and all their clothes. They are certainly graceful; they have music of bells – the sound of all of them playing together is like an organ. These mummers show a thousand graces like these day and night. At night they make shadows of various shapes, like beneditos in Portugal.

The footnote on this passage in the 1944 London edition of Tomé Pires has thrown off a few scholars since this explains beneditos as follows:

The word is perhaps related to the sambenito or sanbenito (saccus benedictus), formerly worn by penitents. It is possible that the tapih, or petticoat, worn by the graceful Javanese dancers, suggested the comparison to Pires.

The thinking here is that Pires is comparing the “various shapes” to beneditos (petticoats), which definition is unknown. But this does not seem correct as the Javanese shadow shapes cannot be compared to petticoats by any stretch of the imagination. Plus, it seems the comparison made in this last sentence by Pires is on the performances of shadow plays, and not the puppets’ shapes. So it is my belief he is really comparing the shadow plays of Java, called wayang, to shadow plays called beneditos, known in Portugal at this time. The passage can also be read in this manner. This makes sense as the Portuguese had possessions in Southeast Asia in the 16th century, when Tomé Pires wrote his book, and the Portuguese could have learned their own play with shadows (called beneditos?) from the Southeast Asians themselves. In addition, the fact that the European shadow plays were always opaque shadows, as in Java, and not colored transparencies as in China, further indicates that the ombres chinoises were derived from the Javanese and not the Chinese.

An explanation of the name ombres chinoises indicating Chinese rather than Javanese derivation may be explained by the fact that the Javanese actually look like the Chinese, at least to illiterate Europeans of the time, when knowledge of other parts of the world was limited. A European seeing a Javanese playing his (or her) wayang may have thought, with such limited knowledge of other cultures and geography, that the dalang (puppeteer) was a Chinese person, hence the popular name ombres chinoises (Chinese shadows).

Henryk Jurkowski, on page 209 in his book referenced above, has speculated that during the Middle Ages, in the mystery plays, shadow figures were known and used. His evidence for this are the following instructions for the stage manager in the 1663 Polish mystery play Utarczka krwawie wojujacego Boga (Bloody Encounter of God Militant): “Here you will show the flagellation through a cloth”; “Here you will show the coronation with the crown of thorns by means of umbrae”; “Here you will show the carrying of the cross by means of umbrae.” Jurkowski says further that

The use of the Latin word umbra for “shadow” suggests that such theatrical practice was nothing new, and almost certainly happened in the earlier mystery plays.



Home Page