Comment on the Malay word count, 1952
M.J. Manning, a non-Malay student of the language, published this critical commentary in 1956 evaluating J. Le Prevost’s Malay Word Count, 1952 (1953), a frequency survey of written Malay compiled by 22 teachers who each examined one million words. Manning’s central argument is that while the survey’s total vocabulary of 3,854 distinct words is a useful cross-section, the individual frequency counts are unreliable due to the inconsistent and unstandardised material each counter selected, and he proposes a restructured ten-category methodology that would yield genuinely comparable data.
Summary
Manning approaches the Malay Word Count from the perspective of the foreign pupil rather than the native scholar, and his commentary is organised around two axes: the product (the word lists themselves) and the process (the survey’s design). On the product side, he is struck by the small total of 3,854 distinct words recorded from 22 million words examined, and by the dissonance between the list and established pedagogical resources such as Hamilton’s Easy Malay Vocabulary. He identifies specific anomalies that undermine confidence in the frequency figures: sultan is credited with only one occurrence per million while raja scores 757; hijau (green) appears just once despite being flora’s predominant colour; and bendahara, a term abundant in Malay historical literature, is counted at only two occurrences while the lesser title laksamana reaches 56. Manning attributes these discrepancies to the practice of treating some words as proper nouns and to the varying literary specialisations of individual counters.
On the process side, Manning’s most damaging observation is a direct comparison of two adjacent counters. Counter No. 15 (allocated initials N, O, Pa–Pau) produced a total frequency of 34,064, while Counter No. 16 (Paw–Pz) produced only 2,338, despite covering a comparable number of distinct words (197 versus 196). Manning demonstrates that No. 16’s list includes common basic words such as pinggan, pisang, pisau, and pokok at counts of one to seven per million, figures he considers implausible. The root cause, he argues, is that each counter selected his own material within broad proportional guidelines (30% newspapers, 35% children’s textbooks, 15% adult non-fiction, 10% adult fiction, 5% correspondence, 5% religious books), with no restriction on period or specific works. A counter who specialised in classical literature would produce radically different frequencies from one who drew on modern newspapers, yet both results were pooled into a single undifferentiated list.
Manning’s constructive proposal is a ten-category survey in which all counters examine identical, precisely specified passages: newspapers (news), modern abstract writing, religious books, recorded colloquial Malay, the “Classics,” modern fiction, textbooks in arts, textbooks in science, pantuns and folk-tales, and personal correspondence—100,000 words each. This would yield ten separate frequency lists, each internally reliable, from which an omnibus list could be derived if needed. He further argues that such a categorised approach is essential for Malay specifically because the language’s register system (Bahasa Dalam, Bahasa Halus, Literary, Colloquial, Bazaar) means that word frequency is inseparable from occasion and style, and because the language is undergoing rapid lexical change through absorption of foreign terms, particularly via the Jawi press.
Key Findings
- The survey examined a total of 22 million words (22 counters × 1 million each) and recorded only 3,854 distinct words, a figure Manning finds surprisingly small given that the Wilkinson and Winstedt dictionaries each contain 9,000–10,000 entries (pp. 37–40).
- Counter No. 15 produced a total frequency count of 34,064 words while Counter No. 16 produced only 2,338, despite listing nearly identical numbers of distinct words (197 vs. 196), exposing the unreliability of the pooled frequency data (pp. 42–43).
- The word sultan is credited with only 1 occurrence per million while raja scores 757 and baginda scores 288, a disparity Manning considers absurd given their semantic equivalence as royal titles (p. 41).
- The colour vocabulary recorded is strikingly sparse: merah 250, hitam 109, kuning 36, puteh 19, biru 11, hijau 1, kelabu 1, leading Manning to question whether the selection was representative (p. 42).
- The prescribed material proportions for each counter were: newspapers 30%, children’s textbooks 35%, adult non-fiction 15%, adult fiction 10%, personal correspondence 5%, religious books 5% (p. 38).
- Manning proposes a replacement methodology of ten categories of 100,000 words each, including a separate list for recorded colloquial Malay and another for “Modern Abstract Writing,” the latter a category he coins for the pedantic philosophical prose then prevalent in the Jawi press (pp. 46–47).
Conclusion
Manning’s definitive takeaway is that the Malay Word Count, 1952 provides a reasonably representative cross-section of the written Malay vocabulary in its total of 3,854 words, but its frequency figures cannot be regarded as a reliable ratio of any million written words because the inconstant ingredients—unstandardised material, uncontrolled period range, and individual counter specialisation—make the pooled data internally contradictory. The work’s real value lies in demonstrating the feasibility of a systematic word-count and in identifying the need for a categorised, repeatable survey that would serve both pedagogical and lexicographical purposes in a language undergoing rapid lexical change.
Context
- Manning’s commentary represents an early application of corpus-based methodology to Malay lexicography, predating modern computational approaches by decades; his proposed ten-category design anticipates the register-based frequency lists that later became standard in language planning.
- The article engages directly with Winstedt’s stylistic classification (Chapter IX of A Malay Grammar) and extends it by arguing for the parallel status of all registers except Bazaar Malay, a position that reflects the mid-century scholarly consensus on Malay sociolinguistics.