The Unresolved
United Nations Security Council Resolution 242, adopted unanimously in November 1967, called for "withdrawal of Israeli armed forces from territories occupied in the recent conflict." The English text says "territories." The French text says "des territoires" — the territories, meaning all of them. The discrepancy was not an error. Lord Caradon, the British ambassador who drafted the resolution, was explicit: the ambiguity was the agreement. Both sides could accept the text because each could read it as affirming their position. Had the drafters resolved the ambiguity — specifying which territories, or saying "all" — one party would have rejected the resolution and it would not have passed. The document worked because it meant two things at once. Disambiguation would not have clarified the resolution. It would have destroyed it.
Henry Kissinger named this practice "constructive ambiguity" during the shuttle diplomacy of the early 1970s, but the method is older than the term. The United States' position on Taiwan has operated through deliberate ambiguity since 1979: acknowledging the Chinese position that Taiwan is part of China without stating agreement. The word "acknowledge" does different work than "recognize" or "accept." The distinction is not imprecision. It is the architecture. Two governments with incompatible claims can operate within the same framework because the framework means different things to each. The ambiguity is not a gap in the agreement. It is the load-bearing member.
In 1801, Thomas Young passed light through two narrow slits and observed an interference pattern on the far wall — alternating bright and dark bands that could only arise if the light passing through one slit interacted with the light passing through the other. In the quantum formulation developed over the next century, a single photon passes through both slits simultaneously. The photon exists in a superposition of two states — left-slit and right-slit — and the interference pattern is the physical consequence of both states being real at once. The pattern is not produced by photons going left plus photons going right. It is produced by photons going both, where "both" is a third state that is not the sum of the first two.
Place a detector at either slit to determine which path the photon takes, and the interference pattern disappears. The photon behaves as though it went through one slit. The which-path information and the interference pattern are complementary in Bohr's precise sense: obtaining one necessarily destroys the other. The superposition carried information — phase relationships between the two paths — that measurement eliminates. Resolution is not clarification. It is data loss.
The genetic code operates on the same principle at the molecular level. Sixty-one codons map to twenty amino acids. The mapping is redundant: most amino acids are encoded by two, three, four, or six different codons. The redundancy is not random. Francis Crick proposed the wobble hypothesis in 1966: the third nucleotide position in each codon pairs loosely with the anticodon on the transfer RNA. The first two positions require strict Watson-Crick pairing. The third position tolerates mismatches — G can pair with U, inosine can pair with U, C, or A. The "wobble" is a controlled relaxation of the pairing rules at a specific position.
The consequence: a point mutation at the third codon position often produces no change in the amino acid. The DNA sequence has changed but the protein is identical. Roughly seventy percent of point mutations at the third position are synonymous. The wobble position absorbs genetic variation without transmitting it to the phenotype.
A strict one-to-one mapping — sixty-one codons, sixty-one amino acids, no redundancy — would be maximally informative. Every mutation would change the output. It would also be maximally fragile. The wobble position trades information density for error tolerance. The ambiguity at the third position is the mechanism by which the code tolerates the mutation rate that DNA replication actually produces.
In Western tonal music, a dominant seventh chord — say, G-B-D-F in the key of C — creates a sensation of tension that demands resolution to the tonic. The tritone interval between B and F is the source: an interval that sits exactly between consonance and dissonance, equidistant from the octave in either direction. The resolution to C major — B rising to C, F falling to E — discharges the tension. But the tension was the experience. The unresolved chord is where the listener leans forward. The resolved chord is where they sit back. Remove the unresolved state and you do not clarify the music. You remove the force that makes it move.
Wagner built entire operas on this principle. The Tristan chord — the opening sonority of Tristan und Isolde — does not resolve for four hours. The harmonic language of the opera keeps approaching resolution and deflecting, creating a sustained state of yearning that IS the dramatic content. The chord is not ambiguous because Wagner could not decide where it should go. It is ambiguous because the ambiguity is the subject.
The Tristan chord's tonal ambiguity — which key is it in? what function does it serve? — has generated over a century of analytical debate. At least five different harmonic analyses have been published. The disagreement is not a failure of analysis. It is evidence that the chord genuinely occupies multiple functional positions simultaneously. Like the quantum superposition, it is not a mixture of definite states awaiting clarification. It is a state that carries something the definite states do not.
The instinct to resolve — to ask "which slit did it really go through," "which territories does it really mean," "which key is the chord really in" — assumes that the resolved state contains strictly more information than the unresolved one. This assumption is false in precisely the cases where ambiguity is doing structural work. The superposition carries phase information that neither basis state contains. The diplomatic text carries an agreement that neither specific reading permits. The wobble position carries robustness that a precise code would lack. The unresolved chord carries tension that resolution discharges. In each case, the unresolved state is not a deficiency awaiting correction. It is a state that exists in its own right, carrying something the resolved states cannot.
Most ambiguity is noise. Most unclear contracts, imprecise measurements, and unresolved questions are simply unclear, imprecise, and unresolved — and resolving them adds information. The cases above are specific. The test is whether disambiguation produces a net gain or a net loss. When a system uses ambiguity as a channel — when the wobble absorbs mutations, when the superposition generates interference, when the unresolved chord creates motion — then resolution is not clarification. It is collapse.