Quintilian, Institutio Oratoria, Book 7
Harold Edgeworth Butler, Ed.

9. I turn to tile discussion of ambiguity, which will be found to have countless species: indeed, in the opinion of certain philosophers, there is not a single word which has not a diversity of meanings. There are, however, very few genera, since ambiguity must occur either in a single word or in a group of words.

[p. 155] [2] Single words give rise to error, when the same noun applies to a number of things or persons (the Greeks call this homonymy): for example, it is uncertain with regard to the word gallus whether it means a cock or a Gaul or a proper name or an emasculated priest of tybele; while Ajax may refer either to the son of Telamon or the son of Oileus. Again, verbs likewise may have different meanings, as, for example, cerno.¹ [3] This ambiguity crops up in many ways, and gives rise to disputes, mole especially in connexion with wills, when two men of the same name claim their freedom or, it may be, an inheritance, or again, when the enquiry turns on the precise nature of the bequest. [4] There is another form of ambiguity where a word has one meaning when entire and another when divided, as, for example, ingenua, armameniam or Corvinum.² The disputes arising from such ambiguities are no more than childish quibbles, but nevertheless the Greeks are in the habit of making them the subject for controversial themes, as, for example, in the notorious case of the αὐλητρίς, when the question is whether it is a hall which has fallen down three times (αὔλη τρίς) or a flute-player who fell down that is to be sold. [5] A third form of ambiguity is caused by the use of compound words; for example, if a man orders his body to be buried in a cultivated spot, and should direct, as is often done, a considerable space of land surrounding his tomb to be taken from the land left to his heirs with a view to preserving his ashes from outrage, an occasion for dispute may be afforded by the question whether the words mean “in a cultivated place” (in culto loco) or “in an uncultivated place” (inculto loco). [6] Thus arises the Greek theme [p. 157] about Leon and Pantaleon, who go to law because the handwriting of a will makes it uncertain whether the testator has left all his property to Leon or his property to Pantaleon.³

Groups of words give rise to more serious ambiguity. Such ambiguity may arise from doubt as to a case, as in the following passage:⁴—

“I say that you, O prince of Aeacus' line,
Rome can o'erthrow.
”

Or it may arise from the arrangement of the words, [7] which makes it doubtful what the exact reference of some word or words may be, more especially when there is a word in the middle of the sentence which may be referred either to what precedes or what follows, as in the line of Virgil⁵ which describes Troilus as

lora tenens tamen,

where it may be disputed whether the poet means that he is still holding the reins, or that, although he holds the reins, he is still dragged along. [8] The controversial theme, “A certain man in his will ordered his heirs to erect statuam auream hastam tenentem,'” turns on a similar ambiguity; for it raises the question whether it is the statue holding the spear which is to be of gold, or whether the spear should be of gold and the statue of some other material. The same result is even more frequently produced by a mistaken inflexion of the voice, as in the line:

quinquaginta uhi erant centum inde occidit Achilles.⁶

[9] It is also often doubtful to which of two antecedents a phrase is to be referred. Hence we get such [p. 159] controversial themes as, “My heir shall be bound to give my wife a hundred pounds of silver according to choice,” where it is left uncertain which of the two is to make the choice.

But in these examples of ambiguity, the first may be remedied by a change of case, the second by separating 1 the words or altering their position, the third by some addition.⁷ [10] Ambiguity resulting from the use of two accusatives may be removed by the substitution of the ablative: for example, Lachetem audivi percussisse Demeam (I heard that Demea struck Laches, or that L. struck D.) may be rendered clear by writing a Lachete percussum Demeam (that D. was struck by L.). There is, however, a natural ambiguity in the ablative case itself, as I pointed out in the first book.⁸ For example, caelo decurrit aperto⁹ leaves it doubtful whether the poet means he hastened down “through the open sky,” or “when the sky was opened for him to pass.” [11] Words may be separated by a breathing space or pause. We may, for instance, say statuam, and then, after a slight pause, add auream hastam, or the pause may come between statuam auream and haslam. The addition referred to above would take the form quod elegerit ipse, where ipse will show that the reference to the heir, or quod elegerit ipsa, making the reference to the wife. In cases where ambiguity is caused by the addition of a word, the difficulty may be eliminated by the removal of a word, as in the sentence nos flentes illos deprehendimus.¹⁰ [12] Where it is doubtful to what a word or phrase refers, and the word or phrase itself is ambiguous, we shall have to alter several words, as, for example, in the sentence, “My heir shall be bound to give him all his own [p. 161] property,” where “his own” is ambiguous. Cicero commits the same fault when he says of Gaius Fannius,¹¹ “He following the instructions of his father-in-law, for whom, because he had not been elected to the college of augurs, he had no great affection, especially as he had given Quintus Scaevola, the younger of his sons-in-law, the preference over himself. .” For over himself may refer either to his father-in-law or to Fannius. [13] Again, another source of ambiguity arises from leaving it doubtful in a written document whether a syllable is long or short. Cato, for example, means one thing in the nominative when its second syllable is short, and another in the dative or ablative when the same syllable is long.¹² There are also a number of other forms of ambiguity which it is unnecessary for me to describe at length.

[14] Further, it is quite unimportant how ambiguity arises or how it is remedied. For it is clear in all cases that two interpretations are possible, and as far as the written or spoken word is concerned, it is equally important for both parties. It is therefore a perfectly futile rule which directs us to endeavour, in connexion with this basis, to turn the word in question to suit our own purpose, since, if this is feasible, there is no ambiguity. [15] In cases of ambiguity the only questions which confront us will be, sometimes which of the two interpretations is most natural, and always which interpretation is most equitable, and what was the intention of the person who wrote or uttered the words. I have, however, given sufficient instructions in the course of my remarks on conjecture and quality, as to the method of treating such questions, whether by the prosecution or the defence.

[p. 163]