Language Deprivation II: Past Babel and the Communal Care of Culture

Part I: Babel

The myth of the existence of a single originary language dates at least to the Biblical story of Babel (Genesis 11:1-9). From very early on, commentators on both this story and that of Adam naming the animals concluded that this first language was Hebrew. For example, the apocryphal book of Jubilees, 12:25-26, has an angel teaching Abraham what it calls this “tongue of the creation.” There are few outliers: some Muslim writers – al-Ṭabarī and al-Ya’qubi – proposed Syriac as the first tongue, as did the twelfth-century Syriac patriarch Michael. And one first-century Jewish philosopher, Philo of Alexandria, includes an astonishing story, unique in the commentarial tradition, about the animals’ own Babel Tower. Once, all animals had spoken one language, until they had the audacity to ask for the immortality they believed the snake already enjoyed: God smote them for their pride, and they fell into mutual incomprehension. Philo does not believe it (“this also, as they say, is a fabulous story”), yet here it is nonetheless.

Christian exegesis on Babel tends concern itself with the plural verbs of Genesis 1:7 (“let us go down {descendamus} and confuse {confundamus}”), which demonstrate, they say, the existence of the Trinity. It often asserts — as did Remigius of Auxerre — that God did nothing new in dividing languages, but rather only divided the already existing category of language into different modes and into different forms of speaking and understanding (see also Peter Comestor): perhaps the actual creation of language, with all that implied about the creation of human reason, could happen only once; or perhaps even creation ex nihilo, of whatever sort, could happen only once; perhaps the divine imprimatur could not be granted to more than one language; or, finally, perhaps exegetes recoiled from imagining God creating a punishment.

But when they think to make an argument about the first language, Christian exegetes tend to agree with Jewish exegetes: at least from the time of Paradise up until the disruption of Babel, language was only Hebrew (eg, pseudo-Clementine Recognitions I,30; Isidore, Etymologies IX,1; and, here standing in for the twelfth century, Andrew of St Victor; see also Dante, pro I,6, and con here). Hebrew was preserved by Heber, hence “Hebrew” (Ranulf Higden and many others), which acquired its proper name only when it first had to be distinguished from other languages (Augustine City of God 16.4). Hebrew would be preserved – and here Christians distinguished themselves – because it was suitable that the language of salvation should first be proclaimed in the language through which death first entered the world (Alcuin of YorkRemigius of AuxerreAngelomus of Luxeuil). (for a thick set of further citations, Christian and Jewish, see Resnick 56-59).

Interestingly, the twelfth-century Maurice of [the Yorkshire Augustinian priory of] Kirkham (h/t) declared that since English had so few case endings, it was the closest language to Hebrew (235). While this error suggests something about the late twelfth-century Yorkshire perception of the relative complexity of English and French – England’s other dominant tongue – I also have to wonder whether Maurice therefore believed English to be the next-closest language to the language of paradise. One is – or I am, anyway – inevitably reminded of Jan van Gorp (d. 1572), who used Herodotus’s Psamtik story to declare the supposedly Phrygian “bekkos” the same word as the Brabantian “becker” (baker), which proved the antiquity, and hence nobility, of the language of Antwerp (this is why, some of his critics claimed, that he took on the absurd name Goropius Becanus; thanks!, but cf). From Phrygian to Flemish, a conclusion that only seems to be sillier than claims about Hebrew.

Back to Babel: the most influential Christian exegete of the Middle Ages had other fish to fry. Like others, Augustine was bothered by the apparent contradiction between Genesis 11 and the several languages spoken by the Noah’s several sons in Genesis 10:5, 20, and 31: how and when did languages actually diversify (and, he might have asked, was the dispersal over the earth a blessing (“be fruitful and multiply”) or a curse?). Augustine’s  Questions on the Heptateuch deals with one of these problems by proposing that the Babel story must be a flashback (in Latin; in French).

But on the topic of the original language, Augustine tends to be agnostic. While his City of God, cited above, does not deviate from the general trends of exegesis, his Literal Interpretation of Genesis allows only that Adam’s language, whatever it might have been (quaecumque autem illa lingua fuerit), could have survived to the present. His anti-Manichaean Genesis commentary proposes that God might have divided light from darkness by speaking Hebrew, Greek, or Latin, or some other language, but, in fact, “all these expressions,” fiat lux and the like, “are adopted to our intellect….For with God there is pure intellect, without the noise and diversity of languages [sine strepitu et diversitate linguarum].”

Part II: (Not Only) Homo Infans

For those who really wanted to know, something more than speculation was needed. There had to be a test, and this test wondered about children, because they routinely demonstrate the transition from speechlessness to language. Isidore of Seville’s Etymologies X.ii.9 correctly derives the Latin word “infans” from a combination of the negating prefix “in” and the present participle “fans” of the deponent verb “fari,” to speak. An infant is a speechless human. But this is only our first age (“homo primae aetatis”). For an infant will learn speech, “quia adhuc fari nescit” (because it does not yet know how to speak), or it will acquire the full power of speech, once its body becomes suitable for speech, as “the expression of speech is small,” because it “does not yet have a full set of teeth” (nondum enim bene ordinatis dentibus).

The first thing to observe is that Isidore characterizes languages as speaking: his verbs are “fari” and “loqui,” verbs that involve the audible voice (several monastic rules, for example, limited or even forbade speech, but in some cases, elaborated “signa loquendi,” signs for speaking, marked as a special form of speech simply by its being allowed, and by their distinction as signa). Furthermore, simply making noise would not have been an adequate proof of being human. The noise has to be determined to be understandable; otherwise, it was as good as silence. “Muteness” is a condition associated with animality – Old French uses “mue beste” often, while Isidore himself distinguishes “pecudibus mutis” (mute livestock) from humans – so that mere noise is effectively the same as silence.

The question is how this language becomes silence. Isidore offers two answers, and in the process two ways of being infans, laying out without resolving the crux of the problem of the origin of language. Either humans have no speech until they learn it, or they have no speech until their body suits itself for its production. In the first conception, language is secondary to us, or it is primary to us as a group, exchanged between us in ongoing acts of teaching and care; in the latter, it belongs to us as much as our teeth do. Such a rooting of language in physical capacity frustrates ideals of disembodiment so common to claims of human rationality, for if language comes out with teeth, perhaps there is no quality that supposedly makes us more than mere bodies, and certainly more than “mere animals.”

For Isidore, as for most thinkers, the homo infans passes its first age and then becomes homo loquens. On this point, I prefer Agamben, for whom infancy is a key critical term (for example, here and here, from The Agamben Dictionary). I cannot endorse his notion that the inexorably alien quality of language in us uniquely thrusts humans into historicity: as Steve Mentz reminded me in comments to my last post, some whale species have “cultural lives,” and as the New York Times just reported, parrots too can have dialects, which they learn while young and then pass on in turn to members of their own group. Psittacus infans.

I am far more sympathetic to Agamben’s insistence that speechlessness is both the necessary condition and the hope of speech. As much as anyone can, Agamben follows Benjamin in arguing that the basic thing language communicates is communication itself [and for my summary of Agamben, I am relying on his commentators rather than his Infancy and History (1978), which has, quite frankly, aged so poorly that I just can’t imagine reading the whole thing]. To be communication, communication must have silence with it; it needs its inbuilt inadequacy. Inadequacy preserves the possibility of communication being something more than a mere back and forth transmission of needs, desires, and aims, something more than what Benjamin called “the bourgeois conception of language.” Thus this inadequacy, figured as a silence within speech, holds open the possible, which we might take as standing for Agamben’s Messianic suspension of the relation between sovereignty and life, and which I prefer to take, perhaps more mundanely, as a preservation of the helplessness within any social encounter, and a preservation of a extra-linguistic referentiality in any communication.

What is always in communication is the “here I am” of speech, a “here I am” whose silence is the preexistent, inescapable vulnerability of having to be somewhere, of needing to be cared for, heard, and to take up attention that might be bestowed elsewhere. This “here I am” is also a “here we are.” Infancy always is within all communication and all community, and all that hope for community. Infancy is always awaiting any attempt to get to the bottom of language, culture, and our civilizations.

Part III: Deprivation and Responsibility

Isidore has nothing to say about practical efforts to resolve these questions, nor, in fact, do many medieval writers. These date back to Herodotus and his tale of Psamtik, a powerful and long-ruling Pharaoh of the twenty-sixth dynasty around whom other equally legendary stories clustered (for example, two first-century encyclopedias, Pliny’s Natural History (XXXVI.19) and Pomponius Mela’s Chorographia (I.48I.56 in English), credit him with building the first labyrinth). Though the Egyptians reputed themselves to be the “oldest nation on earth,” others argued that the honor belongs to the Phrygians: Psamtik (whom Herodotus calls Psammetichus) wanted experimental confirmation. He commanded that two newborns taken from the common people be raised in isolation by a herdsman who was never to speak in their presence. After two years – and here I quote from an English translation of 1584, “both the little brats, sprawling at his feete, and stretching forth their handds, cryed thus: Beccos, Beccos,” which Psamtik and his advisers understood as the Phrygian word for bread. Later commentators have tended to misunderstand the importance of the story’s punchline: it is less about the origin of language than it is about ethnos: “Language,” as Margaret Thomas explains, “only entered into his plan through his assumption that he could identify the oldest people on the basis of linguistic evidence” (here): first people rather than first language.

The story had its doubters along with its misreaders, Herodotus himself included, who numbers it among the “foolish tales” repeated by the Greeks. Scholars of our own era have observed (through what perhaps may be circular reasoning) that the experimental method seems more Greek than Egyptian, and, in misguided quibble, that the word “Beccos” sounds Egyptian, not like the (mostly lost) Phrygian tongue. Modern professionals in early childhood development and linguistics – but also, dismayingly, some cultural historians – sometimes take the story literally: they trouble to dispute the validity of its design, and flaunt their conscience by condemning Psamtik’s cruelty (for some treatments, herehere (“a surprising story, if true”); “supposedly conducted“; “utterly preposterous experiment“; an “oddity of history“; disapproval of this “peculiar brand of child abuse“; and an amusing delineation of Psamtik’s logical errors, including a failure to distinguish between logos and glossa).

Medieval Europe probably didn’t know the story. Herodotus would get no Latin translation until the middle of the fifteenth century, while his Psamtik story slides into European vernaculars only with Pedro Mexia’s widely popular 1540 Silva de varia lección (and from thence, among other routes, to Claude Gruget’s 1552 French translationhere in English, from 1571). Even by first century of our era, Herodotus tended to be cited, by Cicero among others, only through intermediaries (see Félix Racine here). Until the Renaissance, there is little evidence that the story had any readers, first or second hand, despite its being from the first few books of the history, accessible even to the lazy, the harried, and the only pretentiously learned. This remains the pattern.

Quintilian may be rare exception, although he never quite sustained a a regular interest among medieval readers. He explains that “all language” [or “all speech“; omnem sermonem] comes to us by hearing:

Hence infants brought up, at the command of princes, by dumb nurses and in solitude, were destitute of the faculty of speech, though they are said to have uttered some unconnected words. (Institutes X.1)

Yet the plural “princes” and likewise plural “words” (unless he is counting Bekkos twice) suggest that even Quintilian either got the story second-hand or, less likely, that he knew of yet another ancient experiment.

Two brief allusions survive in early Christian writing. Clement of Alexandria’s Exhortation to the Greeks (the Protrepticus) argues that even if “the Phrygians are shown to be the most ancient people by the goats of the fable,” neither they nor the Arcadians nor the Egyptians nor whatever ethnos we claim to be predate the divine logos, responsible for all of us. Clement’s universalist argument (cf Galatians 3:28 and Colossians 3:11) stumbles through a highly compressed allusion not to Herodotus, but rather to an ancient, confused or confusing commentary on an allusion to Herodotus in Aristophanes’s Clouds. The commentary reads that if “goats nursed the children” – which somehow transforms a human herdsman into a female goat – “it is no wonder that hearing the goat they imitated her voice, and it is a coincidence that such an expression occurs among the Phrygians.” No later reader rescues Clement from this muddle: surviving medieval manuscripts of the Protepticus are exceedingly scarce, and so far as I know, it finds no Latin translator until 1551. Even now, Protepticus tends to be Clement’s least-studied work.

Around the same time, in the last decade of the second century, Tertullian’s To the Heathens (Ad Nationes) tells the story at greater length, in an argument that begins by demanding that the Romans explain what they mean by describing Christians as a “third race” (distinct from the polytheism of Romans and Greeks, and distinct as well from the Jews; this phrase, sometimes used by Christians themselves, sometimes used as an insult against them, conceptually overlaps with the idea of a “third gender“). With what can only be associative logic, Jerome then retells Herodotus’s story of Psamtik, which he probably acquired from a historical compilation (Racine 209; perhaps Varro’s lost Antiquitates rerum humanarum et divinarum): Tertullian focuses on an alternate version, only quickly summarized in Herodotus, in which it’s not a herdsman who tends the children, but a nurse whose tongue has been amputated. This is the field of battle Tertullian settles on: no one could possibly survive this wound, the removal of “the very organ of the breath of life”! Therefore – one can imagine him spitting triumphantly – the story must be false. He has nothing to say about the origin of language itself. Only one medieval manuscript of this work survives, a ninth-century copy owned by none other than the great polemicist Agobard of Lyons; but I know of no medieval quotation of or even allusion to Tertullian’s retelling of story, nor would it appear again until 1625, long after Herodotus and Psamtik made their way back into European writing.

The next historical account of the story appears an astonishing 1700 years after Herodotus, in the thirteenth-century chronicle of the Franciscan historian Salimbene di Adam, who, several decades after the events he claims to be recording, explains that Frederick II, Holy Roman Emperor, wanted to know what language children would spontaneously produce if they were never spoken to, or even dandled (blandirentur; in Latinmodern English translation here). Would it be Greek, Latin, or Arabic, or perhaps their parental language, the kind that can be acquired without training, as if – although Salimbene does not say this – one’s particular ethnic language, of whatever sort, sprang from children naturally, and that therefore there were no universal, foundational tongue? This, says Salimbene, is what Frederick wanted to learn. Instead, he learned this: without affection, babies die [since “non enim vivere possent sine aplausu et gestu et letitia faciei et blanditiis baiularum et nutricum suarum”: one paraphrase here; for a brief treatment of twelfth- and thirteenth-century Latin thinking about early child care, Mary Martin McLaughlin].

Salimbene includes his story amid a set of the emperor’s other enormities. He had a scribe’s hand cut off for spelling his name “Fredericus” instead of his preferred “Fridericus”; he had a man sealed and drowned in a winecask to demonstrate that the soul dies with the body (a point Salimbene counters with a flurry of scriptural citations); and he ordered one of his men go hunting, and the other to sleep through the day, and when the hunter returned, had them both cut open to see who had better digested his food: Salimbene meant Frederick to be understood as a monster. The language deprivation experiment is just one more example of tyranny, not a historical fact in any simple sense.

It is harder to imagine a context for the final example from medieval Europe, which comes from a sixteenth-century Scottish historian, Robert Lindsey, who reports that James IV (d. 1513) had a mute woman raise two children on Inchkeith, a barren island of the Firth of Forth, North of Edinburgh. It is short enough to be quoted in full:

The king also caused tak ane dumb voman, and pat her in Inchkeith, and gave hir tuo bairnes with hir, and gart furnisch hir in all necessares thingis perteeaning to thair nourischment, desiring heirby to knaw quhat languages they had when they came to the aige of perfyte speech. Some sayes they spake guid Hebrew, but I knaw not by authoris rehearse, etc

That final, frustrating “etc” suggests that Lindsey may have had more to say about the matter. But this is it. Later, Walter Scott’s Tales of a Grandfather would scoff at Lindsey: “it is more likely they would scream like their dumb nurse, or bleat like goats and sheep on the island.”

And this, combined with my blog post on Akbar and his afterlife, provides a complete record of language deprivation experiments recorded in histories, from Herodotus to the seventeenth century, in other words, 2500 years. Nearly all follow four key points from the pattern laid down by Herodotus.

  • While the children may be raised in an isolated hut or a fortified, well-guarded house, no account imagines the children denied adequate food, shelter, or clothing, only Salimbene de Adam’s account of Frederick II imagines the children denied emotional care: these stories are not presentiments of the several modern, horrific cases of children suffering in appalling confinement for years on end, alone with an uncommunicative, or crueler parent or caregiver; for the most part, all the children are deprived of is spoken communication with adults.
  • Next, the experiment always takes place out of sight of the potentate: practical reasons demand this (the children cannot hear speech, and the potentate and his court need to speak), but it also ensures that in those rare cases where the children produce spoken language, the potentate never directly witnesses language’s emergence: the mystery of whatever truth is sought (whether ethnic, linguistic, or religious) is always screened from the direct observation of the party that first concocted the experiment.
  • All involve multiple children (between two and thirty, but never just one).
  • Finally, from the perspective of the potentate, the experiment is generally a failure: the attempt to find the true origins of an ethnos or language, or an infantile (and therefore “spontaneous”) proof of the superiority of a given faith almost never occurs.

These attempts do not seek to return to the first conditions of creation. To my knowledge, stories of the creation of humans tend not to imagine humans created as infants, or even as a crowd. The Genesis tradition certainly does not begin in silence, but with (in the second creation story, in Genesis 2) God’s supervision as Adam bestows names on animals, most of whom are given no chance to speak back, and certainly no chance to name themselves.

Instead, these experiments attempt to recapitulate an experience common to all humans, all of whom begin in infancy. This experiment is both a grand experimental investigation into the early days of the human species, following the common metaphorization of history as an individual lifecycle of birth, maturity, and decline, and an attempted understanding of the origins of any given individual, who know without remembering that they acquired language some time after coming into existence.

Yet these are not experiments with individuals. That the experiment is conducted with a crowd of forcibly speechless children suggests two key contradictions at the heart of the deprivation experiment. First, they seek to discover proof that language spontaneously emerges, while also recognizing that language is a cultural product, developed and shared between people. Attempts to discover the “authentic” practice – often considered to be the same as the supposed oldest practice, equally presupposed to have kept its purity across time – tend to want practices that just happen, without predetermination. These hunts for the authentic are therefore hunts for culture that inherently distrust the secondary, considered character of culture: they want a natural culture, as if anything acquired by deliberation, desire, and choice must be inherently suspect. They therefore want the benefits of language, ethnicity, and religion, for example, without having to own up to their choice to live through their particular manifestations of these categories. They want culture without responsibility.

Second, they want the find the origin of language while also recognizing that the language’s purpose is interpersonal communication. Though the experiment wants to know if the authentic language (or ethnicity, or religion) lurks within any given child, it also knows that these practices are always practices of a group: it is impossible to imagine an ethnos or a religious community of one, for example. To put this another way, the experiment (almost always) fails because a practice that has the appearance of being inherent to a group is necessarily an emergent and developmental practice, whose feedback structure frustrates any notion of any single origin.

It is therefore not accidental that these experiments are conducted with children. It’s not simply that children are made to function as historically “prior” to adults, though the connection between child and adult that itself maintains and develops culture means that the child is, culturally speaking, secondary to the adult. It is also that children are considered to be free of culture. To deprive children of the care of culture is thus a chance to wait for culture to emerge without our having to care about it.

No wonder the experiments fail. But of course they are not whole failures. In one instance, sign language emerges, taught by “mute” nurses to the children. In another, the children acquire an ordered voice by imitating the sounds of goats. Communication happens, and it happens without the need for spoken or even a human voice. Connections are made.

And what the reveal, again, is that silence amid language. This is not the silence of failure, an aporia or impossibility at the heart of language, though of course the deprivation experiment might be taken that way. The language deprivation experiment is only partially about language, and those who take it as being about language may be overestimating the absence of silence and vulnerability in their own lives.

My preference is to take the supposed failure of the experiment as evidence of the inescapable persistence of bodies. It is a story whose truth the need for community and care. It is a story that produces another witness to the fact that the transformation of silence into any kind of voice requires someone to take the trouble to listen, to talk back, to acknowledge in any way that this too, in this moment, is someone who needs our help. What the potentate sees, finally, is what he should have known all along: his own helplessness, which cannot be overcome.