The Voice we Assign

My students could not see past the greedy Jew.

For three acts, Shylock was exactly what four centuries of reputation had prepared them to find: the moneylender, the pound of flesh, the villain the comedy exists to defeat. They were not being careless. They were reading the play the way the plot invites you to read it, and the plot is not on Shylock's side.

What changed their minds was not a lecture about prejudice. It was form.

We went looking at how Shakespeare had built the man, and found a combination of instruments he gives no one else in that play. Asides, which peel a character away from the scene and turn him briefly toward us. Direct address, which makes the audience complicit rather than merely present. And monologue: sustained, uninterrupted speech, the dramatic form that says this person has an interior and you are going to spend time inside it. None of these devices argues that Shylock is right. They do something more radical and more quiet. They compel attention. They confer a voice on a figure that Shakespeare's London had every convention available for dismissing without one.

Once my students saw the machinery, they could not unsee the man. Something had changed. And the question that surfaced, 'are we supposed to feel sorry for him?' is, I have come to think, the question that matters most in a young reader's formation, because the play refuses to answer it. It hands over the discomfort and requires you to decide.  

That refusal is not a flaw in the writing. It is the mechanism by which literature does something no argument can.

Societies do not come easily to knowledge of their own cruelties. The record here is depressingly consistent: denial is rarely the last response to atrocity, and very often the first. Communities minimise, reframe, forget; the sociologist Stanley Cohen spent a career documenting the elaborate machinery by which cultures refuse to know what they know. This operates not only at the scale of massacre but at the ordinary scale of exclusion, of silencing, of the crushing roles a society assigns and then blames people for occupying. The knowledge is available. The admission is intolerable. Between the two lies a gap that argument has never managed to close, because argument raises exactly the defences it needs to breach.

Fiction closes it by other means, and the means is a kind of licensed trespass. A story carries immunity; it is, after all, only a story, and that is not the form's weakness but its delivery system. Because nobody is being accused and no policy is on the table, the reader's guard comes down, and the pattern sets in. The Russian critic Viktor Shklovsky called the underlying operation defamiliarisation: art's function of making the stone feel stony again, of returning perception to things that habit has worn smooth. And nothing is more habituated, more worn smooth by daily use, than a social arrangement. Venice cannot see its own economy of contempt; it is the water Venice swims in. Poured into one moneylender who bleeds when pricked, that economy becomes visible, and has stayed visible for four hundred years.  

The dosage is calibrated, and the calibration is craft. Shakespeare hands Shylock a speech pivoting from the demands of the plot to reveal a startlingly vulnerable side of Shylock. Ibsen gives Nora a tarantella, the doll dancing frantically for her life while her husband admires the performance, before giving her the door. Aeschylus keeps Cassandra silent on stage for scene after scene before she erupts into prophecy that remains unintelligible to the Chorus, so that the audience who knows precisely what she knows experiences being disbelieved from the inside. These are not happy accidents of composition. They are engineered breaches in the wall a culture builds against its own reflection.  

That the wall is real can be measured by how hard societies fight the breach. Tartuffe was suppressed for five years; Maus was pulled from an American school curriculum in 2022. Ibsen was pressured into writing an alternative ending for German theatres in which Nora stays; the door slam was literally unbearable to the society it described. The Merchant of Venice has not sat comfortably as a comedy for a very long time.  The immunity is therefore unstable; occasionally, the structure recognises its own contradictions and moves to contain them through censorship. But the censor's response measures exactly how much was getting through. Each case measures the same quantity: the distance between what a culture knows and what it can bear to admit. Literature does not abolish that distance. It builds a bridge rated for exactly as much weight as a reader can carry across.

Which raises the question of what writers actually are, institutionally speaking.

Each of us possesses a faculty for watching our own thinking, catching ourselves mid-error, and noticing a habit as a habit. Psychologists call it metacognition, and treat its absence as a serious impairment: a mind that acts but cannot observe itself acting. A society needs the same faculty and has no natural organ for it. The press reports events. Courts adjudicate cases. Historians arrive, by definition, afterwards. None of these can show a culture the shape of its own behaviour, the standing patterns, the exclusions so old they read as nature rather than choice. That work has fallen, across literate cultures, to their writers and poets. Shelley's claim that poets are the unacknowledged legislators of the world overstates the power and understates the function. Writers do not legislate. They audit. They are the process by which a society entertains second thoughts about itself.

And the work demands its own braveheartedness: the writer must first bear the seeing, the pattern a culture has silently agreed to leave unnamed, before building the figure that makes others see it too.

The audit findings prove unusually durable, because what these figures carry is not period detail but recurring configuration. Cassandra models what philosophy only got around to naming in 2007, when Miranda Fricker defined testimonial injustice: a deficit of credibility attached to the speaker rather than to the speech. Every modern instance of the discounted witness, the ignored whistleblower, the woman disbelieved until corroboration arrives from a more acceptable source, runs on the mechanism Aeschylus staged in 458 BCE. Shylock remains the working diagram of the manufactured scapegoat: a community that engineers a position, forces people into it, and then despises them for occupying it. Nora anatomises every role that is comfortable, honoured, and unsurvivable. The literature holds the master drawings. Fresh instances merely plug in.

Which is precisely why the whole apparatus is more fragile than it looks.

An instrument requires a player. A mirror requires someone willing to stand in front of it. Every one of these carefully constructed figures is completed in a reader or not at all, and readers are manufactured, young, through thousands of hours of a specific cognitive practice that neuroscience has begun to describe with some precision. Maryanne Wolf's work on the reading brain draws the essential distinction: decoding is not deep reading. The deep-reading circuit, sustained attention, inference, the simulation of other minds, the tolerance of ambiguity across hundreds of pages, is not innate. It is built by practice and dismantled by disuse.

A reader who has never stayed with a difficult consciousness for two hundred pages has simply not practised staying with difficult consciousnesses. What is at risk is not empathy in the sentimental sense. It is the trained capacity to be reached, to have one's defences breached by a constructed witness. So the ardent request this author makes is for this community to read and, even better, write.

References

Primary Texts

Aeschylus. Agamemnon. 458 BCE. Translated by Robert Fagles, Penguin Classics, 1977.

Ibsen, Henrik. A Doll's House. 1879. Translated by Michael Meyer, Methuen Drama, 1985.

Molière. Tartuffe. 1664. Translated by Richard Wilbur, Harcourt Brace, 1963.

Shakespeare, William. The Merchant of Venice. c. 1600. Edited by M.M. Mahood, Cambridge University Press, 2003.

Spiegelman, Art. Maus: A Survivor's Tale. Pantheon Books, 1986–1991.

Secondary Sources

Cohen, Stanley. States of Denial: Knowing about Atrocities and Suffering. Polity Press, 2001.

Fricker, Miranda. Epistemic Injustice: Power and the Ethics of Knowing. Oxford University Press, 2007.

Shelley, Percy Bysshe. "A Defence of Poetry." Written 1821; published posthumously in Essays, Letters from Abroad, Translations and Fragments, edited by Mary Shelley, Edward Moxon, 1840.

Shklovsky, Viktor. "Art as Technique." 1917. Translated by Lee T. Lemon and Marion J. Reis, in Russian Formalist Criticism: Four Essays, University of Nebraska Press, 1965.

Wolf, Maryanne. Reader, Come Home: The Reading Brain in a Digital World. Harper, 2018.

The Tongue We Inherit

 In 1977, at the height of an international reputation built on novels written in English, Ngũgĩ wa Thiong'o stopped.

He would write his fiction in Gikuyu from then on. He knew the cost — a readership assembled over two decades, a place in the syllabi of universities that would not teach him in translation, the ordinary machinery of literary survival. He did it anyway, and explained why in Decolonising the Mind nine years later. Colonialism's most durable weapon, he argued, had never been the gun. It was the classroom. A child taught in the coloniser's tongue learns, without anyone saying so, that her own culture is a thing to be studied rather than a place from which one studies.

That decision — a writer choosing the smaller language and the smaller audience — is the purest form of a question now being asked by governments rather than novelists. What do you do with what was done to you? And can it be undone by policy?

The psychology beneath it is older than the politics. Frantz Fanon, writing in 1952, was not describing laws or curricula but something inside the colonised person: an evaluative apparatus installed by empire and left running after empire departed. To speak a language, he wrote, is to take on a world. The Antillean who mastered French acquired status and a permanent internal division — fluent in the tongue that measured him and found him wanting. This is the argument's foundation and it is worth stating plainly, because it is often lost in the noise: the claim is not that colonial languages are ugly, or that colonisers were cruel, though both may be true. The claim is that a hierarchy of worth was installed in people's heads, that it survives the flag-lowering ceremony, and that removing it requires deliberate action.

What complicates every subsequent argument is that "decolonisation" now names at least four different projects, and they are routinely conflated. There is the symbolic — renaming cities, streets, institutions; removing statues. There is the linguistic — changing the language of instruction, administration, courts. There is the epistemic — asking whose knowledge counts as knowledge, which is the deepest and least visible of the four. And there is the material — land, capital, the actual distribution of what was taken. These have different costs, different beneficiaries, and radically different evidence bases. A country can accomplish the first three entirely and the fourth not at all. Some have.

The historical record is unusually good on this, because the experiment has been run repeatedly.

Ireland is the closest structural analogue to India and the most sobering. After independence, Irish was made compulsory in schools and required for entry to the civil service, with the explicit ambition of restoring the language the English had suppressed. A century on, Irish holds constitutional status, receives substantial state investment, is taught to every child for thirteen years — and is spoken daily by a small minority. The policy did not fail for want of commitment. It failed, arguably, because compulsion attached the language to examination rather than to life. A generation learned to associate their ancestral tongue with a paper they had to pass. That is not a small lesson for anyone designing language policy now.

Tanzania offers the counter-case, and a qualification. Julius Nyerere made Swahili the national language and it worked — genuinely, at primary level, with measurable gains in access and comprehension. But secondary and tertiary education drifted back toward English, and the tension has never resolved. The pattern is consistent across cases: the higher the level of education, the stronger the pull toward the global language. Whatever one thinks of that gravity, it is real, and policy that ignores it tends to produce a two-tier system rather than a decolonised one.

Algeria pursued the most aggressive programme, replacing French across administration and education from the 1970s. The results were contested and the cultural conflict severe. It also disadvantaged Berber speakers — which names the pattern that recurs most reliably and is discussed least: decolonisation almost always privileges one indigenous group over others. There is rarely a single pre-colonial nation waiting underneath. There are many, and the one that captures the state gets to define authenticity.

South Africa since 1994 has eleven official languages, extensive renaming, and a continuing argument — Rhodes Must Fall being its most visible moment — over whether English's dominance is colonial residue or practical necessity in a country where no single indigenous language is universal. The post-Soviet states offer a parallel archive: Kazakhstan's shift from Cyrillic to Latin script is being conducted now, in public, and will be worth studying for decades.

So what does the evidence actually support?

For the linguistic case, something quite strong. Children acquire foundational concepts faster in a language they already think in; the research on early mother-tongue instruction is not seriously disputed. There is also a dignity argument that is more than sentiment: a state that conducts its business in a foreign tongue tells its citizens something about whose competence is real. And the epistemic claim holds — knowledge genuinely is lost when languages die, and the loss is not recoverable by translation.

Against it, three objections that deserve better answers than they usually receive.

The first is the one that ought to trouble anyone who cares about equity, and it comes from an unexpected direction. English functions, in much of the postcolonial world, as a ladder. Remove it from state schools while private and elite institutions quietly retain it, and you have not decolonised anything — you have converted a colonial advantage into a class advantage and made it heritable. In India this argument has been made most forcefully not by nostalgists for empire but by Dalit intellectuals, who have argued that English is emancipatory precisely because it bypasses hierarchies encoded in vernacular usage — the language in which no one can place you by your accent's caste. Chandra Bhan Prasad's proposal for a shrine to "English, the Dalit goddess" was deliberately provocative and entirely serious. Any account of English as pure colonial residue has to answer it.

The second is that decolonisation is unusually easy to convert into majoritarian nationalism. The step from remove what was imposed to restore what is authentic is short, and the second half requires someone to decide whose authenticity qualifies. In multilingual states this reliably manufactures new internal hierarchies — which is why the resistance in Tamil Nadu to Hindi imposition, running since the 1930s and unbroken since, is not a footnote to the Indian debate but its centre. The people who object are not defending English. They are objecting to being decolonised in someone else's language.

The third is that symbolic decolonisation can substitute for the material kind. Renaming is fast, visible, and cheap. Land reform, educational equity, industrial capacity and the redistribution of what was actually extracted are slow, invisible and expensive. There is a serious critique — made from the left, not the right — that the symbolic project functions as displacement: the appearance of redress standing in for redress.

None of this settles the Indian case, and I will not pretend it does. The National Education Policy's promotion of mother-tongue instruction in early schooling rests on defensible pedagogical ground. Whether its implementation amounts to reducing access to English, and whom that would benefit, is genuinely contested, varies enormously by state, and is being argued over now by people with better information than mine. What history offers is not a verdict but a set of questions any such policy has to survive: Does it change what elite children learn, or only what everyone else learns? Which indigenous language gains, and at whose expense? Is the symbolic change accompanied by material change, or standing in for it?

There is one finding in all this I have not been able to stop thinking about, and it is psychological rather than political.

Decolonising movements do not usually peak immediately after independence. They intensify a generation or two later. India renamed most actively in the 1990s and after, not the 1950s. The generation that experienced colonial rule directly was, on the whole, busier with electricity and famine. It is the generation that inherited the story rather than the injury that most urgently wants the signs undone.

Marianne Hirsch called this postmemory — the relationship of the second generation to a trauma they did not live, mediated by story and image rather than recollection. She was writing about the children of Holocaust survivors. But the mechanism scales, and a nation is not immune to it. Inherited grievance behaves differently from the remembered kind. It is less specific, less negotiable, and harder to satisfy, because the thing that would satisfy it — the actual experience being undone — is unavailable. What remains available is the symbol. So the symbol carries the whole weight.

That is not a criticism. It is a description, and it explains something the policy debate cannot: why these arguments are so much angrier than their material stakes would predict, and why winning them so rarely settles anything.

Which returns me, finally, to the writers, who have been having a better version of this argument for sixty years.

Chinua Achebe kept English and changed it — "the English language will be able to carry the weight of my African experience, but it will have to be a new English, still in full communion with its ancestral home but altered to suit new African surroundings." Ngũgĩ rejected it and paid. Salman Rushdie argued the formerly colonised could not use the language as the British had but should remake it for their own purposes. And Derek Walcott made the division itself the subject, asking how to choose "between this Africa and the English tongue I love."

Four writers, one inheritance, four irreconcilable answers — and not one of them wrong. What the literature knows, and the policy debate keeps forgetting, is that this was never a problem with a solution. It is a condition with responses. Some responses are better than others; none dissolves the thing.

The strengths of these policies and their dangers are not separate matters that might be weighed against one another and balanced. They are the same fact seen from two sides. The impulse to restore what was taken is the impulse that decides who counts as the taker's heir. The desire to give a child her own tongue is the desire that may, if it is careless, take away her ladder. There is no version of this that is only good.

What history suggests is not that the project is misguided, but that it is unusually easy to perform and unusually hard to complete — and that the difference between the two is measured not in what gets renamed, but in who can still read the sign.

Works Cited

Fanon, Frantz. Black Skin, White Masks (Peau noire, masques blancs). Trans. Charles Lam Markmann. Grove Press, 1967. First published 1952.

Ngũgĩ wa Thiong'o. Decolonising the Mind: The Politics of Language in African Literature. James Currey / Heinemann, 1986.

Hirsch, Marianne. The Generation of Postmemory: Writing and Visual Culture After the Holocaust. Columbia University Press, 2012.

The writers' debate

Achebe, Chinua. "The African Writer and the English Language." Morning Yet on Creation Day: Essays. Heinemann, 1975. — The "new English… altered to suit new African surroundings" passage originates in this essay, delivered as a paper in 1964.

Rushdie, Salman. "'Commonwealth Literature' Does Not Exist." Imaginary Homelands: Essays and Criticism 1981–1991. Granta, 1991.

Walcott, Derek. "A Far Cry from Africa." In a Green Night: Poems 1948–1960. Jonathan Cape, 1962. — Source of "how choose / Between this Africa and the English tongue I love?"