translation menu

Hermes—God of Translators and Interpreters

The Antiquity of Interpreting:  Distinguishing Fact from Speculation

This paper was originally commissioned by the Chair
of the NYU Translation Studies Department
as part of a book to be published by John Benjamins. 
It was intended to be a more concise and more conventionally academic version of the paper I presented at the NYU Translation 2000 Conference, which you can find by clicking
  This entire book project was later abandoned by the Chair.

What follows is an attempt to present the evidence that interpreting was relatively common during prehistory and that interpreters had a fairly important role to play during that long and necessarily undocumented era, though differing in some respects from the role they play today.  By interpreting I define those acts of serving as oral intermediary either between two separate languages or between two less than perfectly comprehensible variants of the same language.  From the outset I shall do my best to distinguish probable facts from speculations based on those facts and to demonstrate precisely how I have arrived at these speculations about an era that is otherwise quite closed to us.  Although one occasionally encounters semi-serious references to interpreting as the "second oldest profession," the following account endeavors to provide what may prove to be a reasonable basis for showing that such an assertion might just possibly be justified.

It should be obvious from the outset that interpreting is necessarilly the prehistory of translation—in fact the ancient Romans and Greeks made no major differentiation between the words "interpreter" and "translator."  Where no knowledge of written languages existed, the only way to mediate between two language forms was necessarily interpreting.  It also ought to be obvious that interpreting had to exist during prehistory—the main obstacle to this thesis lies in the nature of modern scholarly procedure, which requires written records and footnotes to bolster any major claim, when clearly neither can exist for the earliest days of interpreting.  I shall nonetheless present three different kinds of evidence combining to suggest the validity of  this statement.

1.  Verbal evidence springing from the earliest words for "translating" and "interpreting."

2.  Inferential evidence pointing towards a continuity in language processes during both prehistory and history.

3.  Perhaps most remarkably, surviving evidence from the past, to be found in preliterate social groups today, replicating to some extent the conditions prevelant in prehistoric social groups.

Examples of verbal evidence are fairly abundant.  Perhaps the most persuasive is the ancient Greek word hermêneus, which the Liddell-Scott lexicon renders as both "interpreter" and "translator."  But these are by no means the only two translations provided by this most authoritative reference work on Greek studies.  Here are some of the other possible translations presented by its editors:


Interpreter, especially of foreign terms


court interpreter 

matrimonial agent 




Furthermore, Liddell-Scott provides the following possible choices for the related verb form:


interpret foreign tongues




put into words 



write about  (Note 1)

From the outset it is useful to reflect on why so many other possible definitions of both the noun and verb forms—aside from “translate” and “interpret”—are listed.  It is of course no secret that the word "interpret" has two major meanings even in modern usage: interpreting the address of a foreign ambassador as opposed to interpreting the meaning of a trend, an omen, a change in a friend's personal behavior.  But the many meanings provided by Liddell-Scott for the noun, especially "go-between," "broker," and "matrimonial agent," would seem at first glance to open up entirely new territory.  And as we shall see, the word "dragoman" has its own ramifications, indeed its own history.  It is by explaining how these subsidiary meanings—or at least what appear at first glance to be subsidiary meanings—connect with the "main" meaning of hermêneus that the first glimpse into prehistory may become feasible.

Furthermore, there is one other crucial clue connected with the Greek use of this word that is simply not present in the way we use the word "interpreter" in English.  As Lidell-Scott, along with Plato and other ancient authors make more than clear, the Greek word hermêneus is intimately tied up with several of the attributes of the Greek god Hermes. Basically, what the ancient Greeks were saying when they used this verb, hermêneuo, hermêneueis, hermêneuei could be duplicated in English only if our verb for to translate or interpret went "I hermese, you hermese, he or she hermeses." Or more collloquially, "I make like Hermes, you make like Hermese, etc."  In the author's opinion, we are under an obligation to ask one simple, searching question: precisely what is going on here in terms of actual meaning?

Several historical and cultural prerequisites must be satisfied before an adequate answer may emerge.  First of all, it is necessary to visualize the full nature of the divinity of Hermes among the Greeks and how they understood his attributes and duties.   For the Greeks, Hermes epitomized quick-wittedness, ready improvisation, and the deft management of clever solutions.  This alone goes towards explaining his connection to language and interpreting.  Among other duties he also acted as divine messenger, presided over commerce and travel (both clearly linked to translation), and was the tutelary god of all the arts and crafts, including magic and matrimonial match-making.  It can perhaps be forgiven if he was also regarded as the god of thieves and deceit, since such attributes may spring somewhat naturally from some of his other functions.  But precisely how do all these traits fit together into a single image of the interpreter during both historic and prehistoric times?

Here's what Plato had to say about Hermes in Cratylus, one of his principal dialogues on language:

SOCRATES: Well, the name ‘Hermes’ seems to have something to do with speech: he is an interpreter (hermêneus), a messenger, a thief and a deceiver in words, a wheeler-dealer—and all these activities involve the power of speech.
(Note 2)

If this seems too negative a view of interpreting, Plato expresses a more positive view in his dialogue Theaetetus, where he also  provides us with a hint of the role played by interpreters in  ancient Athens.  Attempting to distinguish knowledge from  perception, Socrates teasingly asks Theaetetus whether people truly  know a foreign language merely by seeing it in writing or hearing it spoken. In a reply praised by Socrates, Theaetetus states that we can only know what its letters look like and what its spoken form sounds like...

but we do not perceive through sight or hearing, and we do not know, what the grammarians and interpreters teach about them.  (Note 3)

If anything, Plato grants interpreters a slight advantage over grammarians in this passage, for he assumes that grammarians can only be of use in describing the letters or written form of the language (and of the two ancient Greek words for grammarians, both closely related to the word for "letters," the one Plato uses here is the more demeaning one, usually meaning merely a "schoolmaster,"), while only interpreters can tell us what is truly being said.

It has often been observed that the myths of a culture can record or encode very real historical or even prehistorical knowledge, and the Greek linkage of the god Hermes to the act of interpreting can perhaps be regarded as an instance of this.  It is also important to recall how international and multicultural this figure of a trickster god who creates language truly is and how widely recorded it is in the world's mythology.  Hermes was of course known in ancient Egypt as Thoth, but as Lewis Hyde points out in his book Trickster Makes This World, he can also be found in various guises as the African Eshu, as any number of figures such as Coyote or Raven in Native American folklore, as Loki among the Norse, the child Krishna in Indian tradition, or even China's Monkey King, (and in the latter case we have an example of a tale about a god being inspired by the travels of a real-life translator, the seventh century Xuanzang, immortalized in the Ming Dynasty novel Journey to the West.)  (Note 4)

Now let's take a look at the Latin word for translator and interpreter.  At first glance it looks quite familiar: interpres. But as soon as we look at the definitions provided by Charlton Lewis' standard An Elementary Latin Dictionary, we rediscover much of the same terrain previously covered by "hermêneus:"


 [derived from inter, meaning between, and pres, 
a form of  prehendo, prendo, meaning 
to catch, lay hold of, grasp, take 
Literally: Caught in between]

A middle man




Interpres divum 






(Note 5)

Mercury is of course the Roman name for Hermes, and almost all the definitions found for the Greek word recur for the Latin one, except for "matrimonial agent," (perhaps a custom that had died out by Roman times).

The final word for "interpreter" to be examined is not only the most ancient of the three but has also been in continuous use in various forms over at least the last five thousand years and exists in English even today.  As we have seen, Liddell-Scott provided it as one possible English translation for hermêneus:


Spanish: Trujamán, Triujimán

French: Trucheman, Truchement 

Latin: Dragumannus 

Greek: Dragoumanos 

Arabic: Targuman 

Aramaic: Turgemana 

Mishnaic Hebrew: Targum 

Akkadian: Targumanu 

Its meaning:

about 50% interpreter 

about 50% go-between, mediator, 

middle-man, broker (Note 6)

This word alone takes us back to the very demarcation between written history and the prehistory which must have preceded it.  It is a time that clearly predated even the composition of the Bible, and our tale of Eve being created from the rib of Adam, as Kramer points out, is almost certainly based on the world's first recorded untranslatable pun, involving Hebrew and Sumerian, where the word for "rib," ti, also meant the word for "life-giving." (Note 7)  From such a remote point in time we need only press the slightest bit further into the past to securely locate interpreters as present and actively working during prehistory.

At this juncture our two other forms of evidence also have some light to shed, first among these being simple inference based on the present.  We know for a certainty that the process of recorded history has been one of slow, painstaking amalgamation and growth, often against considerable resistance.  Tribal speech gradually broadened and changed across extended clans, distinctly local dialects merged with larger civic or regional ones or at least yielded to them for purposes of trade and culture, pidgins turned into creoles (at least according to Bickerton), (Note 8) regional speech forms have developed into broader structures shared by entire provinces, and provincial languages have been to some extent dominated by official national languages, each with its own pedigree and history.  And almost equally important during this merging and blending process has been the dogged resistance which accompanied it at every stage.  In fact, many parts of the world, as we are well aware, still have not surrendered to this process. These are areas where tribal, clan, and even family loyalties still remain the most powerful force, where even modern methods of communication have failed to penetrate.

Both this evolution—and its many pockets of resistance—provide at least some idea of what cultural and linguistic developments are likely to have been during prehistory.  If anything, the process must have been even slower and more painful in moving from stage to stage, when it managed to move at all.  Only families, at best extended families or "tribes," are likely to have truly and completely shared a "language" to any sufficient degree, and even here new speech forms could have readily grown up between estranged relatives of the same age group or even among members of the same family separated by a single generation.  Whenever members of such families moved to a new location, they took their form of the language with them to develop independently, necessarily leaving the older variant to do the same.  All forms of modern communication were of course totally absent.  And by the very definition of prehistory, there could be no writing.

This view of the past is also supported by our third form of evidence—while prehistoric societies are behind us, preliterate societies still exist.  They can be found in numerous areas of the globe, but there is no longer any real need to travel to those areas to encounter their inhabitants. The reason for this is that members of these societies have now come to live among us and can often be found only a stone's throw from our secure modern dwellings.  Sometimes possessing only a rudimentary knowledge of the national language spoken in their homeland, they have come legally or otherwise to cities like New York, London, and Amsterdam to seek work and improve their condition. 

And not surprisingly, today's interpreters, especially those working in our hospitals and courtrooms, are often the first to become aware of their presence and the need to deal compassionately with them.  These visitors frequently encounter legal and health problems in our world, since their knowledge of modern medicine and social customs does not always coincide with our own.   In fact, they often arrive on our shores still following their own traditional healing methods and social hierarchies, which may include chieftains, priests, and marriage brokers.

And with those last two words perhaps the circle begins to close, since the phrase "matrimonial agent" appeared in Lidell-Scott as one of the meanings for hermêneus.  As recently as a century ago national governments would still cement peace treaties with a marriage between royal offspring of the two nations.  How much more important the arrangement of such marriages must have been for the countless generations of small families, tribes, and clans of prehistory, desperately competing in many environments for perceived advantages in dwelling places, hunting rights, and the cultivation of crops. 

Both of these methods, inference and the survival of comparable groups today, afford us at least some means of visualizing the lives of these peoples.  Merging these two tools with our knowledge gained from our dictionaries, perhaps we can begin to see how important someone skilled in the crucial nexus of deal-making, marriage brokering, and language skills would have been, and all the definitions provided by the Greek lexicon begin to converge as a single comprehensible entity.

Such skills were not without their dangers for those possessing them.  Returning to the historic era, the following account, taken from Plutarch’s Life of Themistocles, demonstrates just how dangerous being an interpreter could be: 

Themistocles, by the consent of the people, seized upon the interpreter, and put him to death, for presuming to publish the barbarian orders and decrees in the Greek language; this is one of the actions Themistocles is commended for..." 
(Note 9)

Merging some speculation with our facts here, it is probably fair to state that similar events could have happened repeatedly throughout the countless centuries of prehistory, as tribes and clans moved in all directions, and the struggle for survival  remained paramount.  The motives of interpreters could easily be misunderstood, even by their own people, nor would we expect such figures to be enormously popular with members of an opposing group.  Not all interpreters and/or go-betweens succeeeded in their missions, not all attempts at peace-making prevailed, any more than they do in modern times.  In a very real sense, today's interpreter at the UN remains a direct descendant of the followers of Hermes.

At this point it is fair to ask how far back into prehistory we may project the existence of such interpreter-intermediary roles.  Even though it may seem that we are now leaving facts behind entirely and entering the realm of pure speculation, something resembling a lucid reply to this question is nonetheless possible.  We may at least posit the thesis that such linguistic middle-men are likely to have existed as long as there have been human languages, following the impeccable logic that language would have been necessary for human survival?or at least for those families, clans, or tribes that managed to survive.

At this juncture we have come as far as facts—and reasonable projections drawn from them—can bring us.  Any further investigation into the history of interpreting necessarily leads us into the domain of speculation.  Quite simply, as soon as we have allowed that proto-interpreters may have existed coterminously with the history of humanity itself, we have also opened up a hotly debated issue: the origins of language itself. 

 Is it possible that the very earliest of proto-interpreters could have actually witnessed these origins, indeed could possibly have played some form of midwife role in such a process?  It should be obvious that merely by asking such a question we risk entering the domain of "cavemen epics" with scantily clad actresses, concocted "primitive" languages, and anachronistic saurians.  So steeped in speculation is such a realm that no reputable scholar should risk entering it, though as we shall see this has not deterred quite a few highly credentialed scholars from doing so.

Let there be no mistake concerning how long a period we may be discussing here—archaeologists and paleoanthropologists agree that it may measure, depending on our definition of humanity, anywhere between two hundred thousand and four million years.  Not only do books on this subject abound, but an entire scholarly organization, the Language Origins Society, has held annual conferences on this topic since 1985.  Based on the amassed literature surrounding this subject, it would appear that representatives from almost every discipline—including many quite distant from language—have not hesitated to come forward with their theories.

Some recent views have concentrated on the shape of the vocal chords over time, some have built grandiose theories on infinitely small details of neuroscience, others have focussed on so-called logical languages.  Most recently an Oxford anthropologist has suggested that human language came about quite suddenly and all at once a mere 10,000 years ago as the result of a single mutation in our genes.  Other authorities insist that it can only have occurred far earlier when  we developed a larynx capable of producing the many sounds of modern language and a hyoid bone ideally situated to support it.  And many of these experts assume that there had to be some truly momentous event, some great divide, some magical, decisive, and defining moment in which human language suddenly took flight and completely separated itself from those horribly rude and base noises made by animals.

Other papers on this topic, all supposedly aimed at discovering the origins of language, have been concerned with communication in the womb, gesture as proto-language, proto-indo-european root forms, Gestalt psychology, the possible influence of bird songs, paleolaryngeology, echolocation, Chomskyan linguistics, and assorted symbolist, postmodernist, and other literary approaches. Some of these papers have managed to avoid the topic of language origins altogether. (Note 10)  So what is about to be suggested here by a humble historian of translation is perhaps no more speculative than much of what has gone before. 

This author is emboldened by the views of Jacques Guy, (Note 11) an authority on Austronesian languages, who has recently pointed out that there is no reason why language had to develop all at once nor even why it had to possess its full panoply of vowels and consonants from the very beginning.   In his words:

..the lowering of the larynx, supposed to have brought the phonetic range necessary for language, is humbug.  Our much pooh-poohed cousins, Homo neanderthalensis, could have made do with just two consonants (a cough and a retch for instance), two vowels (an aarrrgh and a hey), and eight tones, which assuming the simplest syllable structure, left them way ahead, with 48 different syllables, when Rotokas (a sound-impoverished  Papuan language of New Guinea) manages only 35 with its six consonants, five vowels, and no tones.

Such observations provide at least some measure of support for a theory on language origins, possibly connecting proto-interpreters to this process, discussed by the present author in two previous publications. (Note 12) What follows is an excerpt from the first of these:

As human beings we frequently congratulate ourselves as the only species to have evolved true language, leaving to one side the rudimentary sounds of other creatures or the dance motions of bees.  It may just be that we have been missing something.

On countless occasions TV nature programs have treated us to the sight of various sleek, furry, or spiny creatures busily spraying the foliage or tree trunks around them with their own personal scent.  And we have also heard omniscient narrators inform us that the purpose of this spray is to mark the creature's territory against competitors, fend off predators, and/or attract mates.  And we have also seen the face-offs, battles, retreats, and matings that these spray marks have incited.

In an evolutionary perspective covering all species and ranging through millions of years, it has been abundantly shown time and time again—as tails recede, stomachs develop second and third chambers, and reproduction methods proliferate—that a function working in one way for one species may come to work quite differently in another.  Is it really too absurd to suggest that over a period of a few million years the spraying mechanism common to so many mammals, employing relatively small posterior muscles and little brain power, may have wandered off and found its place within a single species, which chose to use larger muscles located in the head and lungs, guiding them with a vast portion of its brain?

Thus, language may turn out to be something we have created not as a mere generation or nation, not even as a species, but in Von Baer's sense as an entire evolutionary phylogeny.

Such a theory obviously raises as many problems as it purports to solve, and some of these are addressed to some extent in the original treatment.  In the absence of any conclusive evidence from so remote an era, it must be clearly labeled as speculative in nature, though perhaps no more so than some other solutions proposed. Recent animal studies, particularly those in ornithology, suggest that sounds and other signals emitted for a specific purpose even among the same species may vary from band to band. (Note 13)  Such differences could lead to confusion whenever two different bands were obliged to meet or merge for an extended period.  Perhaps early hominids encountered comparable difficulties, and it could be that by resolving them that the earliest proto-interpreters may have arisen.

As promised in our introduction, we have presented an account of interpreting during prehistory, limiting ourselves to the factual record in so far as it is available, and extending our treatment into more or less reasonable speculations only where the factual record is missing and where others have not hesitated to speculate as well.  Springing from such a perspective, it is hoped that at least something resembling a reasonable treatment can now fill a noticeable lacuna in the history of translation and interpreting.


(1) Liddell-Scott, 1964/1889, 315; Liddell-Scott, 1961/1843, 690.

(2) Cratylus, 43 (Translated by C.D.C. Reeve).

(3) Theaetetus, 83 (Translated by H.N. Fowler).

(4) Hyde, 9, 17-38, & passim.

(5) Lewis, 436.

(6) Many widely available dictionaries, including the Houghton Mifflin American Heritage Dictionary of the English Language  provide this derivation. In our first two examples, as we have seen, the same word served for both "interpreter" and "translator," but this was apparently not the case for the Sumerians. From both Black and Meissner & Von Soden we learn that an additional word, sepiru, was in use to denote a "translator-scribe." 

(7) Kramer, 149.

(8) Bickerton, 1—135.

(9)  Plutarch, 18-19. (Translated by Bernadotte Perrin)

(10)  The Language Origins Society published the four volumes referenced below (listed under Wind and von Raffler), and its on-line presence provides at any time access to papers from one to three of their conferences.  It has in more recent years been succeeded by EVOLANG, an organization with similar aims.

(11)  Guy, personal correspondence and THES piece referenced above.

(12) Gross, 1993a, 9-10 & 1993b, 254-255.

(13) Welty and Baptista, 234-38, on bird song dialects.  This is a standard university text on ornithology.


Bickterton, D.  1981.  Roots of Language.  Ann Arbor: Karoma.

Black, J, George, A., and Postgate, N. 1999. A Concise Dictionary of Akkadian. Wiesbaden: Harrassowitz.

Gross, Alex. 1993a.  MT and Language: Conflicting Technologies? Ariadne's Endless Thread.  Sci-Tech Translation Journal, October: 8-10, 16.  Poughkeepsie, NY: American Translators Association. 

———1993b.  Selected Elements from a Theory of Fractal Linguistics. In Scientific and Technical Translation, ATA Scholarly Monograph Series, Vol. VI, Wright, S.E., and Wright, L. (eds), 225-263.  Amsterdam/Philadelphia: John Benjamins.

Guy, J. 1973.  A grammar of the northern dialect of Sakao, a Melanesian language of Espiritu Santo (New Hebrides) .  Canberra: Australian National University.

Guy, J.  2000. It Was the Larynx, Guv. (review of The Origins of Complex Language by Andrew Carstairs-McCarthy) Times Higher Education Supplement, February 11: 29.

Hyde, L. 1998.  Trickster Makes this World.  New York: Farrar, Straus, & Giroux.

Kramer, Samuel.  1963. The Sumerians: Their History, Culture, and Character. Chicago: University of Chicago.

Lewis, C.  1918.  An Elementary Latin Dictionary. New York: American Book Company.

Liddell, H.G. and Scott, R.  1961/1843. A Greek-English Lexicon.  Oxford: Clarendon; 1964/1989.  An Intermediate Greek-English Lexicon.  Oxford: Clarendon.

Meissner, B. and Von Soden, W.  1966. Akkadisches Handwörterbuch.  Wiesbaden: Harrassowitz.

Plato.  1998.  Cratylus. Translated, with Introduction and Notes, by C.D.C. Reeve.  Indianapolis/Cambridge: Hackett.

Plato.  1921/1966.  Theaetetus.  [Loeb Classical Library] Cambridge, Mass.: Harvard University.

Plutarch. 1914/1997. Lives: Life of Themistocles. [Loeb Classical Library] Cambridge, Mass.: Harvard University.

Von Raffler-Engel, W., Wind, J. and Jonker, A. (eds) 1991. Studies in Language Origins, II. Amsterdam: John Benjamins.

Welty, J.C. and Baptista, L. 1988. The Life of Birds.  Fort Worth: Saunders College and Harcourt Brace Jovanovich.

Wind, J., Pulleyblank, E.G., de Grolier, E., and Bichakjian, B.H.  (eds) 1989. Studies in Language Origins, I. Amsterdam: John Benjamins, 1989.

Wind, J., Jonker, A.,  and Rolfe, R. (eds) 1994 Studies in Language Origins, III. Amsterdam: John Benjamins.

Wind, J., Bichakjian, B.H., Nocentini, A.,and Chiarelli, B. (eds) 1992. Language Origin: A Multidisciplinary Approach. (Papers from the joint 1988 LOS Meeting and the NATO Advanced Study Institute on Language Origin). Dordrecht, The Netherlands: Kluwer.

Wu, Chen-En and Kherdian, D. 2000. Monkey: A Journey to the West.  New York: Random House.

translation menu
linguistics menu
to top