Issue #47 December 2021

Socrates’ Garden of Letters: On Writing, Relays, and Discourse

Jean Arp - "Objects Placed on Three Planes Like Writing" - (1928)

Émile Benveniste’s general linguistic theory was largely based on vocal speech acts. He only turned to writing in a sustained manner at the end of his career, in a series of lectures given at the Collège de France during the 1968-9 academic year. Due to a stroke suffered late in 1969, these lectures were never formalised or fleshed out. 

The penultimate session used Plato’s critique of writing from Phaedrus, though only as part of a demonstration showing how writing had been undervalued and viewed inconsistently in Indo-European history. The ultimate session sketched out the central relationship between writing and language, notably that both “signified in the exact same way.” “Language,” for the moment, can be considered a vocally-realised system of signs shared by a community. While Benveniste did not directly argue against Plato—the reference illustrated a difference of views between Ancient Greek, German and Old Icelandic—, his conclusion appeared to undermine Socrates’ position by denying writing’s uniqueness. 

The final lesson ended with, “We are at the beginning of a reinterpretation of many concepts (all those that touch on language). The very notion of ‘language’ must be larger; it must include more notions than what we attribute to it.” We will never know all he had in mind, though we can speculate—the role Benveniste attributed to philosophy—based on the lessons that he intended to keep community discourse, or dialogue, as a core concept of language. This essay dives deeper into Phaedrus to draw out implications of including writing in discourse. Using the examples of John Dewey’s “Great Community” and Benedict Anderson’s “Imagined Community,” it also briefly touches on how the implications can play out in contemporary society.


Benveniste’s discussion of pronouns, adverbs and the like from “The Nature of Pronouns” (1956) gives a sense of his approach to discourse prior to the final lessons. The intent of the article was to show how, unlike nouns and verbs, the form and function of pronouns were not unified:

“Some belong to the syntax of the language, others are characteristic of what we call “instances of discourse,” which is to say discrete and individually unique acts by which language is realised in speech by a locutor.”

From the article’s opening paragraph, two aspects of language were set. Syntax covered the general rules for putting elements of a language together in sentences. Instances of discourse escaped from the general rules and instead relied on traits of a specific act. Sometimes pronouns refer back to nouns in a sentence or set of sentences, regardless of the broader context. Sometimes they refer to that context:

“I is the “individual who utters the current instance of discourse containing the linguistic I.” As a result, in introducing the situation of “allocution,” one obtains a symmetrical definition for you, as the “individual spoken to in the current instance containing the linguistic instance you.” These definitions single out I and you as a category of language and link to their position in language. One does not need to consider the specific forms of the category in the given languages, and it matters little if the forms must explicitly appear in the discourse or can stay implicit.”

Anchoring to an instance of discourse also affects demonstratives, such as “this” and “that,” and adverbs like “here” and “now.” Benveniste started with personal pronouns because the person speaking is at the centre of this context. They are the implicit or explicit “I” of the act, their placement in space and time determines the here and now, their accompanying gesture allows us to know which book this or that book is, and what recipient of the words is the “you.”

It is telling that the “I” in the article spoke and did not write. Plato teased out a relevant difficulty of writing near the end of Phaedrus:

“Socrates: And now may I be allowed to ask you a question: Would a husbandman, who is a man of sense, take the seeds, which he values and which he wishes to bear fruit, and in sober seriousness plant them during the heat of summer, in some garden of Adonis, that he may rejoice when he sees them in eight days appearing in beauty? at least he would do so, if at all, only for the sake of amusement and pastime. But when he is in earnest he sows in fitting soil, and practises husbandry, and is satisfied if in eight months the seeds which he has sown arrive at perfection?

Phaedrus: Yes, Socrates, that will be his way when he is in earnest; he will do the other, as you say, only in play.

Socrates: And can we suppose that he who knows the just and good and honourable has less understanding, than the husbandman, about his own seeds?

Phaedrus: Certainly not.

Socrates: Then he will not seriously incline to ‘write’ his thoughts ‘in water’ with pen and ink, sowing words which can neither speak for themselves nor teach the truth adequately to others?

Phaedrus: No, that is not likely.

Socrates: No, that is not likely—in the garden of letters he will sow and plant, but only for the sake of recreation and amusement; he will write them down as memorials to be treasured against the forgetfulness of old age, by himself, or by any other old man who is treading the same path. He will rejoice in beholding their tender growth; and while others are refreshing their souls with banqueting and the like, this will be the pastime in which his days are spent.

Phaedrus: A pastime, Socrates, as noble as the other is ignoble, the pastime of a man who can be amused by serious talk, and can discourse merrily about justice and the like.

Socrates: True, Phaedrus. But nobler far is the serious pursuit of the dialectician, who, finding a congenial soul, by the help of science sows and plants therein words which are able to help themselves and him who planted them, and are not unfruitful, but have in them a seed which others brought up in different soils render immortal, making the possessors of it happy to the utmost extent of human happiness.”

The Socratic language, despite the less-than-ideal circumstances of being read in translation a couple thousand years after the fact, is still quite flowery. Only, the flowers are dead; well-preserved, but dead. They cannot “speak for themselves” or fully “teach the truth” as they have become fixed rhetoric rather than fruitful dialogue.

The issue was more modest for Socrates. Phaedrus was primarily, though far from exclusively, a dialogue about love, sparked by a speech arguing that a non-lover would be a better romantic partner than a lover. The speech was written and then recited to multiple audiences. As such, the words were not adapted to each hearing. This was akin to planting seeds inappropriate for the conditions; okay as a pastime but not really serious.

A couple precisions should be made here about what we mean by “writing.” Socrates mentioned two forms; the first, “in pen and ink,” as noted above, and “another kind of writing etched on the tablets of the mind.” The interest here is with the former, mapping onto what Benveniste called readable iconic (or symbolic) representation. This writing can be read aloud, as the speech in Phaedrus is many times. Being read aloud does not alone make it into a unique verbal utterance.

Written words unable to “speak for themselves” went further than the frivolous word husbandry above. The quote Benveniste used describes the breakdown of dialogue:

“Socrates: I cannot help feeling, Phaedrus, that writing is unfortunately like painting; for the creations of the painter have the attitude of life, and yet if you ask them a question they preserve a solemn silence. And the same may be said of speeches. You would imagine that they had intelligence, but if you want to know anything and put a question to one of them, the speaker always gives one unvarying answer. And when they have been once written down they are tumbled about anywhere among those who may or may not understand them, and know not to whom they should reply, to whom not: and, if they are maltreated or abused, they have no parent to protect them; and they cannot protect or defend themselves.”

Let us segue into Benveniste’s use of the passage. The similarity between writing (graphē), painting (graphō) and attitude of life (zōgraphia) is lost in translation. The detail, raised by Benveniste, indicated a different approach from the “runa” versus “bok” development in some other Indo-European languages. Those in turn could be compared to the “scribō,” “meljan,” “pisati,” etc. forms. Languages in the family had a chaotic constellation of forms to express writing, revealing a weak writing tradition relative to families to the east and south. In Phaedrus, the weakness could also be glimpsed when Socrates brought up Theuth (Thoth), the Egyptian god of, among other things, writing, something missing in the Greek pantheon.

Benveniste’s next and final step was to show that, across, and even in the most chaotic, language families, the expression—the naming of the operation—of writing was paired with that of reading. To this, he added certain aspects of the phenomenon of writing from earlier lessons, notably:

  • that it uses a limited set of icons to express practically unlimited ideas. This differentiated writing, which was read, from pictograms, which were understood. 
  • that the set referred back to the language rather than independently signifying a concept. “Language [orally expressed] and writing signify in exactly the same way.” 

These premises allowed him to conclude that writing in practise was a unique intersubjective linguistic act between writer and reader akin to the speaker-listener pairing. It was effectively a “relay” of speech.

Even with the argument only sketched out in a series of lectures, it was too expansive for this essay. We certainly do little justice to the etymological and pictographic study supporting it and have said nothing about the importance of writing for linguistic analysis and self-semiotisation. Nonetheless, it is clear Benveniste saw language as both a semiotic system and “instances of discourse.” Discourse was effectively an I-you dialogue whose significance cascaded from pronouns through demonstratives, adverbs, adverbial phrases and onward. Writing was a visually iconic form of this discourse.

On one level, to shift back the broader implications of the quote, Plato would agree with Benveniste’s conclusion. The words of both the pre-written and extemporaneous speech were seeds destined for some audience. Both could communicate truth, persuade and do everything a speech was intended to do. Ultimately, both could flourish as ideas in the community. Non-vocalised writing was more ambiguous, due to the Greek language grouping writing, drawing, painting, scratching and so on together and to Socrates describing memory as a sort of writing. As any writing can be vocalised, the ambiguity is interesting but not pivotal.

The issue of Socrates’ garden is the multiplication of instances for one discourse. The initial instance comes when the text is written; the seed is created. The “I” is set, with all the initial linguistic context, at time of writing. A more or less defined “you” is also fixed, ranging from the named recipient of a personal message to the potentially vast and varied audiences of the rhetorician. Following instances come with each reading or reciting. 

Jean Arp "Six White Forms and One Gray Make A Constellation On A Blue Ground" - (1953)

In many cases, multiple instances might not matter. Socrates suggested planting the seeds for “recreation and amusement” as such a case. Benveniste pointed out that some texts fall more under the “syntax of language” than “instance of discourse”:

“One can imagine an extensive linguistic text—a scientific treatise for example—where I and you do not appear a single time; inversely it would be difficult to conceive of a short spoken text where they are not employed.”

The relay would still require a broader understanding of “instance of discourse” as part of the larger notion of language. Specifically, how do we consider situations where the “you” is either not well—or at all—defined, implicitly or explicitly, or misses the mark? If language is a shared system, it would need a feedback loop to ensure continued mutual comprehension. Regardless of how viable the seed might be, out of season it could not play its role in the linguistic circle of life.

Benveniste hinted at of this difficulty in earlier lectures:

“Speech [Oral expression] is realised formally in discrete words, one assembles the parts of a whole one after the other, while “writing” is first conceived as totality, it synthetically presents an entire train of ideas, it recounts an entire story. In this sense, “writing” resembles much more “interior language” than a chain of discourse.”

“Interior language” was not etched in the way Socrates described it, but rather “has a global, schematic, non-constructed, non-grammatical character. It is an allusive language”:

“Interior language is rapid, incoherent, because one always understands oneself. It is always a situated language, in the present context, that is part of the condition of language, therefore intelligible to the author and to him alone. But transfer this interior language, conditioned by the relation of a locutor with himself as and in a unique, ever-changing experience and circumstance, into a form intelligible to others and losing in its written aspect the natural relation with the occasion which is that of interior language, is a considerable task.”

This is not a complete picture, since even if the writer succeeds in making the text intelligible to some others in some external circumstances, the text remains the same while both others and circumstances are “ever-changing.” The notion of writing as a totality rather than a link in the chain of discourse nonetheless describes reasonably well the difference Plato saw between rhetoric and dialogue. Moreover, the question of intelligibility was central: “Once written down they are tumbled about anywhere among those who may or may not understand them.”

The conclusion to which Benveniste arrived, then, was not as solid as it might seem. Languages that name writing invariably do so as part of a writing-reading pair, just like speaking and listening. The “I” that writes or speaks does so with a “you” implicitly or explicitly set. The roles are reversed in the reply, and so on and so forth. These acts, taken as a whole, are the realisation of language, a shared (intersubjective) semiotic system in the community. Only, just because the pairing formally exists does not mean it always functions. While Plato’s position on writing was undoubtedly coloured by linguistic and cultural influences—if writing was actually important there would be a local god to champion it—, his description of how the realisation could fall apart continues to be compelling.

For the present-day, the writing-speaking divide is, while remaining important for self-semioticisation and linguistic analysis, fairly useless for our purposes. Benveniste began his first lesson by stating:

“We live in the civilisation of the book, of the book read, of the book written, of writing and of reading. Our thought is constantly, at whatever level, informed by writing.”

In the end, he concluded writing was a “relay.” This is a far more relevant concept, as it can cover the full range of recorded and transmitted linguistic acts. Regardless of whether the act is written or verbal, it is the separation between the “I” and the “you,” with all their contextual particularities, that is key.

This can lead us in many directions, as relays have multiplied in the modern world. Although following any of them would be beyond the scope of this essay, we will briefly mention two. John Dewey, in his 1927 The public and its problems, described the United States as a society whose technical advances outstripped the general capacity to conceptualise and meaningfully communicate them:

“Symbols control sentiment and thought, and the new age has no symbols consonant with its activities. Intellectual instrumentalities or the formation of an organized public are more inadequate than its overt means. The ties which hold men together in action are numerous, tough and subtle. But they are invisible and intangible. We have the physical tools of communication as never before. The thoughts and aspirations congruous with them are not communicated, and hence are not common. Without such communication the public will remain shadowy and formless, seeking spasmodically for itself, but seizing and holding its shadow rather than its substance. Till the Great Society is converted into a Great Community, the Public will remain in eclipse. Communication can alone create a great community. Our Babel is not one of tongues but of the signs and symbols without which shared experience is impossible.”

“The physical tools of communication” had, if anything, broadened linguistic communities more than ever before. However, despite the shared language, shared conceptualised experience was lacking. The public was, in essence, a “you” unable to see itself as such, let alone as a reciprocal “I” in a dialogue. Due to the “Great” nature of society, the linguistic community had perhaps not so much broadened as stretched to a point where “instance of discourse” no longer scaled. Alternatively, community ended at the limit of Benveniste’s “instance” and a different approach would be needed to describe linguistics acts in Dewey’s hypothetical “Great Community.”

Benedict Anderson’s modern nation state-level “imagined communities,” as described in the book of the same name, are at the same scale as Dewey’s “Great” groupings and even more explicitly based on the written word. “The most important thing about language,” in the form of “print capitalism,” “is its capacity for generating imagined communities, building in effect particular solidarites.”

“[A nation] is imagined as a community, because, regardless of the actual inequality and exploitation that may prevail in each, the nation is always conceived as a deep, horizontal comradeship. Ultimately it is this fraternity that makes it possible, over the past two centuries, for so many millions of people, not so much to kill, as willingly to die for such limited imaginings.”

The writer-reader pairing was brought up repeatedly. As but one example, Anderson’s description of Filipino writer José Rizal’s novel Noli Me Tangere included: 

“The casual progression of this house from the ‘interior’ time of the novel to the ‘exterior’ time of the [Manila] reader’s everyday life gives a hypnotic confirmation of the solidity of a single community, embracing characters, author and readers, moving onward through calendrical time.”

Through novels and newspapers, in increasingly standardised languages taught in increasingly standardised schools, the “I” and the “you” became a national “we.” Sometimes the “we” was delimited by language, sometimes by colonial administrative areas or other considerations, but “print,” later with “allies,” “especially radio and television,” was always key. While most community members would never meet each other, they could imagine each other as fellow readers and, later, spectators.

Anderson’s position was, however, ambiguous. The Rizal example above was supposed to show burgeoning Filipino consciousness, yet it was limited in important ways to Manila. Elsewhere, he noted print material was aimed at the “reading classes,” such as “courtiers and ecclesiastics, rising middle strata of plebeian lower officials, professionals, and commercial and industrial bourgeoisies.” On one hand, both writer and reader could imagine a broader national audience. On the other, the semantic intent of the text’s “you” rarely corresponded to that audience.

There was also little linguistic reciprocity. Imagined community was characterised as “unisonance;” “an experience of simultaneity.” It was a relationship comparable to that evoked by the “imagined sound” of “national anthems” and other “ceremonial poetry.” In a footnote, Anderson asked the reader to “contrast this a capella chorus with the language of everyday life, which is typically experienced decani/contoris-fashion as dialogue and exchange.” This effectively unidirectional and passive bond may have sufficed to motivate people to kill or be killed for their country. It nonetheless shares the shortcomings Plato saw in rhetoric and falls short of Benveniste’s conception of the writing-reading pairing.


Benveniste’s last lectures, published only in 2012, were pulled together by Jean-Claude Coquet and Irène Fenoglio using Benveniste’s notes supplemented by those of linguists in the audience. These days, courses at the Collège de France are recorded and disseminated through a variety of channels. The reference to Phaedrus was from audience notes, so was likely not a core idea Benveniste intended to express. And, as mentioned in the introduction, he did not have the chance to further develop any of it.

The quote, among other anti-writing arguments Socrates made in the dialogue, was pertinent to what Benveniste was saying and where he might have gone with an expanded notion of language. Writing is unique, though not, strictly speaking, as an instance of discourse. As an instance, it is simply one example of an expressive act that is generally separated from the communicative or receptive act. While there are many forms of writing and speaking, the distance between writing and reading has historically been greater than that of speaking and listening. Just as today, the distance associated with relayed expressive acts is greater than direct ones. While the point seems obvious, the writing-reading pairing, perhaps not unlike “the physical tools of communication” and democratic institutions, tends to obscure the fact that the “you” of discourse is not so obvious.

Trent Portigal is a writer of eclectic curiosities. Novels include Our New Neolithic Age (2021), Simulated Hysteria (2020), Death Train of Provincetown (2019) and The Amoeba-Ox Continuum (2017).


December 2021


The Ethics of Proximity: A Defense of Different Ethical Duties to Friends and Family

by Antonio Wolf

On Kierkegaard and the Autobiographical Self

by Timofei Gerber

Socrates' Garden of Letters: On Writing, Relays, and Discourse

by Trent Portigal

The Specter of Reality TV

by Alexandra Fall