between the two in vocative constructions and negation. Thus they say yā Zaydu and yā ʿAmru (“O Zayd” and “O ʿAmr”), with a final u-vowel, as in His words, “O Adam” and “O Noah.” With nouns in genitive constructions and similar forms, they say yā ʿAbda Allāhi (“O servant of Allah”) and yā ghulāma Zaydin (“O Zayd’s servant”), as in His words: “O children of Adam,” “O children of Israel,” “O People of the Scripture,” “O people of Yathrib,” and “O our people, respond to Allah’s caller.” These and similar examples have the first noun of the genitive construction in the accusative. The same applies to compounds formed by fusion: khamsa (“five”) is not the same as khamsata ʿashara (“fifteen”); rather, combining the words changes the meaning.

That being so, anyone who said, “‘Five’ is literal when it means five, but figurative in ‘fifteen,’” would be ignorant. For the latter expression is not the former, even though the word “five” occurs in both places: it has entered into a different construction. This type of construction is established by linguistic convention, just as the genitive construction itself is.

Likewise, the expressions “the wing of travel,” “the wing of humility,” and “the back of the road” constitute a different construction, in which the noun is annexed to something other than that to which it was annexed in the other instance. This is neither like the word standing alone, such as “five,” nor like its being joined to another word, as in “twenty-five.”

This same point may be made in the fourth argument: whether an assignment of meaning prior to usage is established, or whether “assignment” means what is known from usage, on either assumption this expression in a genitive construction was assigned and used only for this meaning. Nothing else is understood from it; indeed, it admits no other meaning. Understanding what is intended by it requires only a contextual indication of meaning other than what has been mentioned