Similar Words - Different Meanings

This page helps you distinguish between meanings of pairs of similar words used in COMP9414/9814 Artificial Intelligence, or pairs of words, one of which is familiar from some other context, and the other of which is used in a technical sense in COMP9414/9814.

WordsWhat's the difference?

The word agreement refers to a particular feature (often written agr) that is used to indicate whether the particular form of a noun or verb or determiner being used is singular or plural. In languages other than English, other parts of speech (such as adjectives) may have singular/plural forms, and may thus become involved in the agreement distinction. As well as singular/plural, agreement distinguishes (in English) between 1st, 2nd, and 3rd person, that is, between pronoun and noun forms like I, we (first person), you (second person), he, she, it, and they (third person). The word "agreement" comes from "agree" - the idea being that the parts of the sentence must agree with each other on whether the sentence is about a singular grammatical subject. Thus we goes is not OK, because we says that the subject is plural, while goes is a singular form of the word go - we and goes do not agree in this sense.

The word argument is a general mathematical term that is heavily used in both Prolog programming and Natural Language Processing terminology. From a Prolog perspective, an argument is one of the comma-separated objects that appears between a pair of parentheses, after a predicate symbol (also known as a relation name). Thus, in likes(mary, pizza), you can see that mary is the first argument, pizza is the second argument, and likes is the relation name.

The word augment is used to indicate the use of grammar rules with extra information added to them in the form of feature restrictions. It just happens that agreement, described above, is a feature commonly used to in augmented grammar rules. Other features used include the sem feature, used to record the logical form of a constituent of a sentence (such as a noun, a noun phrase, or the whole sentence), the var feature, and the subcat feature. "Augment" means to increase something - thus the feature information in augmented grammar rules increases the rules, in size, in information content, and also in terms of power to describe the grammar and semantics of the language.


The word determinant you are most likely to have met in a technical context in mathematics, where it is a value associated with a matrix (or array) of numbers. If the determinant is non-zero, then the matrix has an inverse, for example.

The word determiner is not related to determinant. It is the name for a part of speech, that is, a class of words, that includes the words a, an, some, the, this, that, these, those, etc., that help to determine exactly what the noun phrase that they begin refers to. For example, this pizza would refer to a pizza near to the speaker, while that pizza would refer to a pizza further away from the speaker.


The word infinite (related to infinity) is used in mathematics to describe a set of objects, often numbers, but could be vectors or points, etc., that is so large that in cannot be counted, even in principle. So the set of all whole numbers, like 1, 234, –86, and so on, is infinite.

The word infinitive is used in linguistics to refer to a form of a verb, or a way of using a verb in a sentence. In English, the infinitive form of a verb starts with the word to and is normally followed by the base form of the verb: to be, to eat, to have, etc. In other European languages, the infinitive form is likely to be a single word: être, manger, avoir in French, for example. Verb phrases can also be infinitive (by including an infinitive verb form) - to be a hero, to eat one's dinner, to have a heart attack. It is not the case that the presence of the word to means that an infinitive will follow, as to is also used as a preposition, so it can also be followed by a noun phrase: e.g. [go] to the bank, [give it] to his mother.

When used in the infinitive context, to is sometimes referred to as the infinitive particle. When used in a prepositional phrase, to is of course a preposition.

To complicate things slightly, sometimes infinitive constructions have extra bits. The extra bits will be auxiliary verbs, adverbs, and not (the negative particle). Examples include: to really like (with the adverb really in the middle), to have eaten (auxiliary have in the middle), and to not know (negative particle in the middle). Constructions like this are called split infinitives if the extra bit is an adverb or the word not.

under construction
under construction
under construction

Copyright © Bill Wilson, 2008.
Bill Wilson's contact info

UNSW's CRICOS Provider No. is 00098G