A Study of the “Natural” Equilibrium between the Sentence and Clause Length

Applying Menzerath–Altmann's Law to Translated Texts in Japanese

Abstract

In order to understand balancing the lengths of linguistic elements contained in a sentence, we conducted a survey using the Menzerath–Altmann's law (MAL), which is often used in quantitative linguistics research in Europe and Canada. Using texts from two editions of the Japanese translation of a French book, Le Petit Prince, as our study material, we construct a dataset to create a function with the length of the sentence measured by the number of clauses as x, the length of the clauses that constitute the sentence measured by the number of morphemes as y, and the number of characters as y'. The regression analyses were performed on the linguistic length dataset, employing the MAL equation. As Altmann stated, the association between the sentence length and clause length is expressed by a decreasing function. The MAL holds even in translated texts. The regression curves of the two editions tended to be similar to each other. However, some data deviated from the MAL decreasing function, specifically, for cases with additions, longwinded explanations, parenthetic expressions, or words expressing thoughts without quotations. It can be regarded that these cases are intended to have an expressive effect by means of breaking the balance of length among linguistic elements, which is typical in Japanese expressions.