Argument

The Babel Argument

Intro

There are more than 7,000 languages in the world. They sound nothing alike. Some have no word for "left" and "right" and use compass directions instead. Some make you say how you know what you are saying. And yet a Bible translator, a diplomat or a bilingual child can take what was said in one and say it in another. It may take more words, and some music is lost, but the meaning gets across.

Even dead languages give up their meaning. For centuries no one could read Egyptian hieroglyphs. Then, in 1822, Jean-François Champollion used the Rosetta Stone to crack them, and the old priests' words spoke again. What they meant had been there all along, waiting to be found.

That points to something remarkable: all these languages are different roads into one shared country of meaning, a country that was there before any of the roads were built. Meaning is made of thoughts, things that are about other things and can be true or false. A whole realm of thoughts that belongs to no human language looks like the thinking of a Mind. The Gospel of John names it: "In the beginning was the Word" (John 1:1, ASV). Babel scattered the languages. The Word underneath them was never scattered.

◎ See this page's graph and discover linked topics →

In full

The Babel Argument is a likelihood argument from the structure of linguistic meaning. Its explanandum is the conjunction of three facts: radical surface diversity among human languages; the near-universal inter-translatability of their cognitive content, including the recovery of meaning from dead scripts; and the universal human capacity to acquire, and even to create, any human language. These facts support the existence of a language-independent realm of propositions, contents that many sentences in many languages can express. A deductive spine (P1 to P6) then argues that propositions bear marks characteristic of thought (intentionality, truth-aptness, logical structure), so that a realm of them existing prior to and independently of every human language is far more expected if there is an eternal Mind whose thoughts they are, the Logos of John 1:1-3, than on naturalism, which must either deny propositions (nominalism), accept them as mind-independent abstracta with unexplained aboutness (godless Platonism), or reduce them to types of human cognitive acts (Soames), which leaves too few of them. The evidence is drawn from comparative linguistics, the history of decipherment, developmental psychology and the philosophy of language. The conclusion is confirmatory, not demonstrative, and it is sensitive to one contested premise, the effability of every proposition in every language, which is defended in a moderate form.

Format note

This is a braided case, the house format for new Ris3n Arguments: a valid spine whose premises each rest on independent strands of evidence, a knot, a likelihood ledger and a defeater gate. It braids five fields. Language contributes the comparative fact that languages differ in what they must say but not in what they can say. Objective truth contributes propositions, the bearers of truth that do not change when the sentence does. Psychology contributes the infant who can learn any language and the deaf children who built one. History contributes decipherment, meaning recovered from scripts no living person could read. Theology contributes the Logos, the Word that was before every word.

The spine

Validity. P1 to P3 establish the data, P4 infers the realm of propositions as their best explanation, P5 says such a realm is far more expected on theism than on naturalism, and P6 turns that into confirmation. The form is an inference to the best explanation (P4) feeding a likelihood comparison (P5) closed by modus ponens on P6.

Strength of the conclusion. Confirmatory. A naturalist may reject P4 (no propositions) or accept propositions and reject P5 (they are mind-independent or human-made). The ledger states what each option costs.

P1, Translatability

Inductive strand

1. Diversity is real. Ethnologue, the standard catalogue, lists more than 7,000 living languages. Their grammars vary enormously in sound, word order, tense, evidential marking and spatial reference.
2. Jakobson's principle. Roman Jakobson, "On Linguistic Aspects of Translation" (in Reuben Brower, ed., On Translation, Harvard, 1959), argued that all cognitive experience can be conveyed in any existing language, and that "languages differ essentially in what they must convey and not in what they may convey." Guy Deutscher, Through the Language Glass (2010), defends the same distinction from a modern linguist's perspective.
3. Katz's effability principle. Jerrold J. Katz ("Effability and Translation," 1978) proposed that every proposition can be expressed by some sentence in every natural language, and drew the consequence that translation between any two languages is possible in principle. A single counterexample would refute it; critics have offered candidates, and the thesis remains debated.
4. Practice. Scripture translation projects have rendered the whole Bible or part of it into thousands of languages, and the translators' working assumption, borne out daily, is that meaning can cross.

Abductive strand

Convergence

Structural linguistics (Jakobson), formal semantics (Katz), modern popular linguistics (Deutscher) and translation practice agree on P1 in its moderate form: content crosses even when form does not. The weak form of linguistic relativity, that grammar nudges attention and habit, is compatible with P1 and is not denied here.

Objections and rebuttals

1. "Quine showed translation is indeterminate." Rebuttal. W. V. O. Quine (Word and Object, 1960) argued that rival translation manuals could fit all the behavioural evidence. But Quine applied the same indeterminacy to understanding one's own language, so it is a thesis about meaning in general, not evidence that some languages say what others cannot. See W.V.O. Quine.
2. "Poetry cannot be translated." Rebuttal. Sound, rhythm and wordplay belong to form, which is exactly what Jakobson said differs. P1 concerns cognitive content. Jakobson himself treated poetry as the limiting case of form.

P2, Recoverability

Inductive strand

1. Egyptian. The Rosetta Stone, a decree of 196 BC in hieroglyphic, Demotic and Greek, gave Jean-François Champollion the key; his Lettre à M. Dacier (1822) announced the decipherment of the hieroglyphs, unread for about fourteen centuries.
2. Linear B. Michael Ventris deciphered the Linear B script in 1952 and, with John Chadwick, showed it wrote an early form of Greek ("Evidence for Greek Dialect in the Mycenaean Archives," Journal of Hellenic Studies 73, 1953).
3. Cuneiform. The decipherment of Akkadian cuneiform in the nineteenth century, aided by the trilingual Behistun inscription, opened Mesopotamian literature, including texts that bear on Genesis. See ANE Legal Codes, Comparative Context.

Abductive strand

Convergence

Egyptology, Aegean archaeology and Assyriology reached P2 independently, with different scripts and different methods. Their common lesson is that meaning is not trapped in the community that first spoke it.

Objections and rebuttals

1. "Some scripts, like Linear A, remain undeciphered." Rebuttal. That shows the limits of evidence, not of meaning. No one thinks Linear A tablets mean nothing; we lack the key.

P3, Universal access

Inductive strand

1. Any infant, any language. Patricia Kuhl, "Early Language Acquisition: Cracking the Speech Code" (Nature Reviews Neuroscience 5, 2004), reviews evidence that young infants can discriminate the speech sounds of all languages before narrowing to their native ones. An adopted child acquires the language of the home, whatever the birth parents spoke.
2. Children who made a language. Ann Senghas, Sotaro Kita and Aslı Özyürek, "Children Creating Core Properties of Language: Evidence from an Emerging Sign Language in Nicaragua" (Science 305, 2004), documented how deaf children brought together in new schools from the late 1970s developed a sign language, with successive cohorts introducing the discrete, combinable structure typical of human languages.
3. The capacity is prior to the language. The same pattern underlies the codex's Argument from the Pre-Given Logos, which develops the poverty-of-the-stimulus case. This page uses only the fact that the capacity is universal and language-neutral.

Abductive strand

Convergence

Developmental psychology (Kuhl), the study of emerging sign languages (Senghas and colleagues) and generative linguistics converge on P3. The evolved-faculty account is accepted as the mechanism; the argument's question is about what the faculty reaches.

Objections and rebuttals

1. "This is just Chomsky again." Rebuttal. No. The Pre-Given Logos argument concerns the innate grammar of the receiver. The Babel Argument concerns the shared content that every grammar can carry, which a language faculty does not by itself explain.

P4, Propositions

Inductive strand

1. Same content, many sentences. "Snow is white," "La neige est blanche" and "Der Schnee ist weiß" say the same thing. What they share is not ink, sound or grammar but content, which philosophers call a proposition.
2. Frege's third realm. Gottlob Frege, "Der Gedanke" (1918), argued that thoughts in this sense are neither physical objects nor private ideas, but a "third realm" that many minds can grasp.
3. Augustine's inner word. Augustine, De Trinitate XV.10.19, distinguished the word in the mind from every spoken word: the inner word is neither Greek nor Latin nor of any other language.

Abductive strand

Convergence

Patristic theology (Augustine), the founder of modern logic (Frege) and contemporary philosophy of language converge on a shared realm of content. See Universals and Anyone Who Affirms Universals Must Affirm God.

Objections and rebuttals

1. "Propositions are a philosopher's fiction." Rebuttal. Then so is "what she said," "the claim you just denied," and "the same thing in French." Ordinary speech quantifies over contents constantly, and translation practice depends on it.

P5, Thought-likeness

Inductive strand

1. Aboutness. Propositions are about things; the proposition that snow is white is about snow. Aboutness, or Intentionality, is the mark of the mental.
2. The argument from propositions. Alvin Plantinga listed an argument from the nature of propositions among his "two dozen or so" theistic arguments, and Lorraine Juliano Keller develops it in "The Argument from Intentionality (or Aboutness): Propositions Supernaturalized" (in Jerry Walls and Trent Dougherty, eds., Two Dozen (or so) Arguments for God, Oxford, 2018).
3. Theistic conceptual realism. Greg Welty, "Theistic Conceptual Realism" (in Paul Gould, ed., Beyond the Control of God? Six Views on the Problem of God and Abstract Objects, Bloomsbury, 2014), argues that propositions are best understood as divine thoughts.
4. Too many for finite minds. Keller presses a "scarcity" point: there are more truths than finite minds will ever think, so if propositions are thoughts, only an infinite mind could supply them.

Abductive strand

Convergence

Philosophy of mind (intentionality as the mark of the mental), philosophy of language (Frege, Soames) and philosophical theology (Augustine, Plantinga, Welty, Keller) agree on the key data: propositions are representational. They disagree on the ground, and the abductive table compares the options.

Objections and rebuttals

1. "This is just Platonism with God added." Rebuttal. It is Platonism with the puzzle removed. Plato's forms do not think; propositions represent. Locating them in a Mind explains why they are about anything at all.
2. "Why the Christian God and not a generic mind?" Rebuttal. The argument confirms a Mind whose thought precedes all languages. John 1:1-3 adds that this Word is personal and creative, and Acts 2 and Revelation 7:9 add that every language is finally gathered back to him. The specificity is theological fit, not deduction.

P6, The likelihood principle

Inductive strand

The likelihood principle is standard in probability theory and in scientific confirmation and is accepted by many critics of theistic arguments.

Abductive strand

Rejecting it would undercut decipherment itself, which proceeds by asking which reading makes the texts most expected.

Convergence

Bayesian and likelihoodist schools agree on this comparative use.

Objections and rebuttals

1. "No prior, no conclusion." Rebuttal. The ledger is prior-agnostic; priors are argued in the Cumulative Case for Christian Theism.

The knot

The strands come from different disciplines and people: comparative linguistics (Jakobson, Katz, Deutscher), decipherment history (Champollion, Ventris and Chadwick), developmental psychology (Kuhl, Senghas and colleagues), and philosophy of language and theology (Frege, Augustine, Plantinga, Keller, Welty). Two overlaps are counted once: Jakobson's principle and Katz's effability express one claim, and P3 partly overlaps the Pre-Given Logos argument, so this page uses only its universality. The decisive step, from a realm of propositions to a Mind, rests on P5 alone and is weighed once.

The ledger

Prior-agnostic: the ledger asks only how expected each datum is under each hypothesis. The ratings are reasoned judgments from the sections above, not measurements.

Net. The linguistic and psychological rows alone are close to even; they establish the data rather than favour theism. The weight sits in the last three rows: once the shared realm of meaning is granted, its thought-like character and its size are far better explained by an eternal Mind. The argument is as strong as P5.

The defeater gate

Naturalized propositions. The strongest objection comes from contemporary naturalistic theories of propositions. Scott Soames (Rethinking Language, Mind, and Meaning, Princeton, 2015) and Jeffrey King (The Nature and Structure of Content, Oxford, 2007) argue that propositions are not eternal abstracta at all but are grounded in the cognitive acts or linguistic facts of real human agents. On Soames's view, a proposition is a type of act of predicating something of something, so its aboutness comes from the agents who perform it. That explains intentionality without God, and it explains why many languages share content: they are used by the same kind of cognizers performing the same kinds of acts.

The reply has three steps.

1. Scarcity. Most truths will never be thought by any human: truths about unobserved stars, about numbers too large to write, about counterfactual histories. If propositions are types of human acts, those truths have no proposition to be. Keller's scarcity objection presses exactly this point.
2. Priority. On the human-act view there were no propositions before there were humans, so it was not true, before humans, that the earth orbited the sun. Most philosophers, and every working scientist, treat such truths as holding long before anyone thought them.
3. What the gate concedes. The human-act theory is the best account of propositions as we grasp them, and the argument welcomes it as a description of our side. It is the right theory of the copy, not of the original. Revise it to include acts no human performs and it becomes the thesis of this page.

Concessions

• Grant: languages shape habits of attention (weak relativity). Now collect: habits of attention are differences of form; the content still crosses, which is what P1 says.
• Grant: an evolved faculty explains how infants learn language. Now collect: it explains the receiver, not the shared content all receivers reach.
• Grant: propositions grasped by humans depend on human minds. Now collect: then truths never grasped by humans depend on a mind that is not human.

Cheatsheet

The 30-second version (read aloud). There are seven thousand languages, and they are wildly different, yet anything you say in one can be said in another. Even dead languages give up their meaning: we read the hieroglyphs again. So every language is a door into one shared house of meaning. That house is made of thoughts, things that are about other things and are true or false. Thoughts come from minds. A house of thoughts older than every language, bigger than every human mind, is what you would expect if "in the beginning was the Word."

Fast facts:

1. Ethnologue: more than 7,000 living languages.
2. Jakobson (1959): languages differ in what they must convey, not in what they may convey.
3. Champollion (1822): the hieroglyphs read again after about fourteen centuries.
4. Senghas, Kita and Özyürek (Science, 2004): deaf children in Nicaragua built a language with core linguistic properties.
5. Keller (Oxford, 2018): there are more truths than finite minds could think.

The closing line. Babel scattered the words; it never scattered the Word.

Cross-examination

Round 1, Can meaning cross?

• Q: Is there anything said in Hebrew that cannot, in principle, be said in English?
• If they say no: then languages share one realm of content. What is that realm?
• If they say yes: ask for an example, and then ask how they know what it means.

Round 2, What is meaning?

• Q: When a Greek sentence and its English translation say the same thing, what is the thing they both say?
• If they say "nothing, just similar sentences": then explain why unrelated languages are similar in exactly the way translation needs.
• If they say "a proposition": then ask what a proposition is about, and how something non-mental can be about anything.

Round 3, Before us

• Q: Was it true that the earth orbited the sun before anyone thought so?
• If they say yes: then there were truths before human minds. Whose thoughts were they?
• If they say no: then astronomy describes what became true when we noticed it.

Dodge table

Live-cite kit

Scripture (ASV):

• John 1:1-3: In the beginning was the Word, and the Word was with God, and the Word was God. The same was in the beginning with God. All things were made through him; and without him was not anything made that hath been made (John 1:1-3, ASV)
• Genesis 11:1: And the whole earth was of one language and of one speech (Genesis 11:1, ASV)
• Genesis 11:7: Come, let us go down, and there confound their language, that they may not understand one another's speech (Genesis 11:7, ASV)
• Acts 2:8, 11: And how hear we, every man in our own language wherein we were born? ... we hear them speaking in our tongues the mighty works of God (Acts 2:8, 11, ASV)
• Revelation 7:9: a great multitude, which no man could number, out of every nation and of all tribes and peoples and tongues, standing before the throne and before the Lamb (Revelation 7:9, ASV)
• Psalm 19:3: There is no speech nor language; Their voice is not heard (Psalm 19:3, ASV)

Scholarly:

• Roman Jakobson, "On Linguistic Aspects of Translation" (1959): "languages differ essentially in what they must convey and not in what they may convey."
• Augustine, De Trinitate XV.10.19: the inner word belongs to no language.
• Lorraine Juliano Keller, "Propositions Supernaturalized" (2018): only an infinite mind supplies enough thoughts to be the propositions.
• Ann Senghas, Sotaro Kita and Aslı Özyürek (Science, 2004): children creating core properties of language.

Aphorism: "Seven thousand doors, one house."

Connection to Scripture

• Genesis 11:1-9, Babel: the languages are divided, but the God who divides them understands them all.
• Acts 2:6-11, Pentecost: the same mighty works of God heard in every hearer's own language, the reversal of Babel without abolishing the languages. See Pentecost.
• John 1:1-3, the Word prior to every word, through whom all things were made. See Logos Christology and logos.
• Revelation 7:9, every tongue gathered before the throne: diversity kept, unity restored.
• Daniel 7:14, all peoples, nations and languages serving the Son of Man.

Scholarship

Published academic works this page draws on, each matched to a publisher, index or library catalog record.

• Reuben Arthur Brower, On translation (Harvard University Press, 1959). Open Library
• Jean-François Champollion, Lettre à M. Dacier (1822). Open Library
• Guy Deutscher, Through the language glass (2010). Open Library
• Franz Guenthner, Meaning and translation (New York University Press, 1978). Open Library
• Jeffrey C. King, The Nature and Structure of Content (Oxford University Press, 2007). doi:10.1093/acprof:oso/9780199226061.001.0001
• Patricia K. Kuhl, "Early language acquisition: cracking the speech code," Nature Reviews Neuroscience 5 (2004). doi:10.1038/nrn1533 PubMed 15496861
• Willard Van Orman Quine, Word and object (1960). Open Library
• Ann Senghas, Sotaro Kita and Asli Özyürek, "Children Creating Core Properties of Language: Evidence from an Emerging Sign Language in Nicaragua," Science 305 (2004). doi:10.1126/science.1100199 PubMed 15375269 open access
• Michael Ventris and John Chadwick, "Evidence for Greek Dialect in the Mycenaean Archives," The Journal of Hellenic Studies 73 (1953). doi:10.2307/628239
• Jerry L. Walls and Trent Dougherty, Two Dozen (or so) Arguments for God (2018). Open Library

See also

• Ris3n Arguments, the hub of ris3n's original arguments
• Arguments, the master index
• Argument from the Pre-Given Logos, the innate capacity for language; this page concerns the shared content every language carries
• Argument from the Naming Convergence, Kripke on naming; this page concerns meaning, not reference
• Anyone Who Affirms Universals Must Affirm God, the wider argument from abstract objects
• Universals, the realism and nominalism debate behind P4
• Intentionality, the aboutness behind P5
• Logos Christology, the doctrine of the Word
• Genesis 11 and Tower of Babel Objection Defeater, the Babel narrative and its critics
• Pentecost, the reversal of Babel
• W.V.O. Quine, the indeterminacy objection
• Cumulative Case for Christian Theism, where this argument contributes