Balanced word pairs

Balanced pairs are the best known of the Quran's numerical claims: opposite or related words said to occur exactly as often as each other. This world and the Hereafter, angels and devils, life and death. Most of the lists in circulation go back to Abd al-Razzaq Nawfal's al-I'jaz al-'Adadi li-l-Qur'an al-Karim, and they are shared as a single striking fact: the Quran keeps its opposites in balance.

This page does two things. It counts the famous pairs, using the method set out on the word-count page. Then it runs a test the famous pairs cannot: thirty opposites chosen before any of them was counted, to see whether opposites in the Quran balance more often than chance would allow.

The famous pairs

Eight well-known pairs were checked on the word-count page and all hold, each under one rule applied to both words: this world and the Hereafter (115), angels and devils (88), man and woman (24), Adam and Jesus (25), benefit and corruption (50), public and openly (16), tongue and admonition (25), and "say!" and "they said" (332). See those counts.

Below are seventeen more that are widely shared. Each was counted every reasonable way: every form of the word, the singular only, without attached pronouns, with the article al- only, and every word from the root. "Count it" opens the counter on the comparison shown.

Two hold under one rule for both words: faith and disbelief, and guidance and mercy, both at the numbers usually quoted. Two more reach their numbers only by counting one word alone and the other as a whole root. In ten, the quoted number is right for one word of the pair, usually as a whole root, but the other word does not match it under any of the five rules. Three do not balance at all.

Some of the misses are by one or two, such as patience at 103 against hardship at 102. A count made by hand from a printed concordance can group a word or two differently from this data. That is a reason to read near misses generously. It is not a reason to treat them as balanced: the claim is exact equality, and on these rules the counts are not equal.

A test named in advance

Every famous pair was found by someone looking for pairs that balance. However many such pairs there are, they cannot tell us whether balance is a property of the Quran's opposites or a selection from among them. That needs pairs chosen before counting.

So we chose thirty. They are opposites and companions the Quran itself sets side by side: night and day, heaven and earth, light and darkness, the Garden and the Fire, truth and falsehood, ease and hardship. They were fixed, with the three counting rules, before any of them was counted. Every pair already famous for balancing was left out, since those were chosen because they balance.

For each pair we also measured chance, shown in the last column. We took every pair of words with the same frequencies, each within 20% of the test words' counts, and counted them the same three ways. The share that balance is how likely a pair of words that common is to balance with nothing arranged at all. Added up over the thirty pairs, it gives the number of balanced pairs to expect by chance.

One pair of the thirty balances: summer and winter, each mentioned once, in the same verse (Quran 106:2). Words that rare balance by chance almost two times in three. Chance predicts 1.7 balanced pairs out of thirty; the Quran's opposites give one. Some come close, such as the Garden at 147 and the Fire at 145, but none of the others is equal under any of the three rules.

What this shows

Ten pairs, the eight on the word-count page and faith/disbelief and guidance/mercy here, balance under a single stated rule, and anyone can check them verse by verse. Those are real features of the text, and the counter lists every verse behind them.

The test shows something the famous lists cannot. Taken as a whole, the Quran's opposites do not occur in equal numbers any more often than chance would produce. The famous pairs are a selection from a text in which most opposites do not balance: night 84 and day 57, heaven 310 and earth 461, truth 247 and falsehood 26. There is no sign that the Quran's opposites are arranged to balance, and nothing is lost by that. The Quran mentions the Garden and the Fire, truth and falsehood, as often as its message needs, and its frequencies follow its meaning rather than a numerical scheme.

The weaker items in the lists, where the rule changes between the two words or the number fits only one of them, are best dropped when the Quran's case is presented. A sceptic will check them, and finding that they fail can cast doubt on the claims that do hold. The Quran's own challenge was never numerical. It is to "produce a surah like it" (Quran 2:23), a challenge about its language and guidance, which this does not touch. The objection is taken up in full in Do numerical and ring-composition miracles prove the Quran divine?

Sources

• Quranic Arabic Corpus, morphology version 0.4, Kais Dukes, University of Leeds: corpus.quran.com, with the corrections at github.com/mustafa0x/quran-morphology
• Abd al-Razzaq Nawfal, al-I'jaz al-'Adadi li-l-Qur'an al-Karim, the source of most of the pairs in circulation
• The test of thirty pairs: scripts/quran/balanced-pairs-test.mjs, with the pairs and rules fixed before counting
• Quran translations: The Clear Quran, Dr. Mustafa Khattab