The first question which arises here is, how many degrees of comparison has the English adjective (and, for that matter, the adjective in other languages, such as Russian, Latin, or German)? If we take, for example, the three forms of an English adjective: large, larger, (the) largest, shall we say that they are, all three of them, degrees of comparison? In that case we ought to term them positive, comparative, and superlative. Or shall we say that only the latter two are degrees of comparison (comparative and superlative), whereas the first (large), does not express any idea of comparison and is therefore not a degree of comparison at all? Both views have found their advocates in grammatical theory. Now, if we define a degree of comparison as a form expressing comparison of one object or objects with another in respect of a certain property, it would seem that the first of the three forms (large) should not be included, as it does not express any comparison. Then we should have only two degrees of comparison larger, (the) largest, and a form standing apart, coinciding with the stem from which the degrees of comparison are formed, and which may be described as the basic form. However, in a very few adjectives the basic form differs from the stem in sound. This difference is of some importance, though it is not reflected in the spelling. This applies to two adjectives in -ng, namely long and young; their stems are [loηg-] and [jAηg-] and the degrees of comparison formed from these stems are, longer ['loηgэ], longest ['loηgist] and younger [jAηgэ], youngest ['jAηgist]. The basic forms, on the other hand, are long [loη] and young [jAη], without the final [-g] which is impossible after [-η] in modern literary English. 1 A somewhat similar phenomenon is found in adjectives ending in -r or -re, such as poor, pure, rare, sure. Their stems are [рuэr-], [pjuэr-], [reэr-], [Suэr-] and the suffixes of the degrees of comparison are added on to these stems, whereas the basic form loses its final [-r], unless it is followed without pause by a word beginning with a vowel, as in the phrases poor idea, rare image, and the like. Now it is well known that not every adjective has degrees of comparison. This may depend on two factors. One of these is not grammatical, but semantic. Since degrees of comparison express a difference of degree in the same property, only those adjectives admit of degrees of comparison which denote properties capable of appearing in different degrees. Thus, it is obvious that, for example, the adjective middle has no degrees of comparison. The same might be said about many other adjectives, such as blind, deaf, dead, etc. However, this should not be taken too absolutely. Occasionally we may meet with such a sentence as this: You cannot be deader than dead. In a novel by E. Hemingway the hero compares the ways one and the same word sounds in different languages: Take dead, mort, muerto, and todt. Todt was the deadest of them all. But as a rule adjectives having such meanings do not appear in forms of comparison A more complex problem in the sphere of degrees of comparison is that of the formations more difficult, (the) most difficult, or more beautiful, (the) most beautiful. The question is this: is more difficult an analytical comparative degree of the adjective difficult? In that case the word more would be an auxiliary word serving to make up that analytical form, and the phrase would belong to the sphere of morphology. Or is more difficult a free phrase, not different in its essential character from the phrase very difficult or somewhat difficult"? In that case the adjective difficult would have no degrees of comparison at all (forming degrees of comparison of this adjective by means of the inflections -er, -est is impossible), and the whole phrase would be a syntactical formation. The traditional view held both by practical and theoretical grammars until recently was that phrases of this type were analytical degrees of comparison. Recently, however, the view has been put forward that they do not essentially differ from phrases of the type very difficult, which, of course, nobody would think of treating as analytical forms. Let us examine the arguments that have been or may be put forward in favour of one and the other view. The view that formations of the type more difficult are analytical degrees of comparison may be supported by the following considerations: (1) The actual meaning of formations like more difficult, (the) most difficult does not differ from that of the degrees of comparison larger, (the) largest. (2) Qualitative adjectives, like difficult, express properties which may be present in different degrees, and therefore they are bound to have degrees of comparison. The argument against such formations being analytical degrees of comparison would run roughly like this. No formation should be interpreted as an analytical form unless there are compelling reasons for it, and if there are considerations contradicting such a view. Now, in this particular case there are such considerations: (1) The words more and most have the same meaning in these phrases as in other phrases in which they may appear, e. g. more time, most people, etc. (2) Alongside of the phrases more difficult, (the) most difficult there are also the phrases less difficult, (the) least difficult, and there seems to be no sufficient reason for treating the two sets of phrases in different ways, saying that more difficult is an analytical form, while less difficult is not. Besides, the very fact that more and less, (the) most and (the) least can equally well combine with difficult, would seem to show that they are free phrases and none of them is an analytical form. The fact that more difficult stands in the same sense relation to difficult as larger to large is of course certain, but it should have no impact on the interpretation of the phrases more difficult, (the) most difficult from a grammatical viewpoint. Taking now a general view of both lines of argument, we can say that, roughly speaking, considerations of meaning tend towards recognising such formations as analytical forms, whereas strictly grammatical considerations lead to the contrary view. It must be left to every student to decide what the way out of this dilemma should be. It seems, on the whole, that the tendency towards making linguistics something like an exact science which we are witnessing to-day should make us prefer the second view, based on strictly grammatical criteria. If that view is adopted the sphere of adjectives having degrees of comparison in Modern English will be very limited: besides the limitations imposed by the meaning of the adjectives (as shown above), there will be the limitation depending on the ability of an adjective to take the suffixes -er and -est. 1 A few adjectives do not, as is well known, form any degrees of comparison by means of inflections. Their degrees of comparison are derived from a different root. These are good, better, best; bad, worse, worst, and a few more. Should these formations be acknowledged as suppletive forms of the adjectives good, bad, etc., or should they not? There seems no valid reason for denying them that status. The relation good: better = large: larger is indeed of the same kind as the relation go: went = live: lived, where nobody has expressed any doubt about went being a suppletive past tense form of the verb go. Thus, it is clear enough that there is every reason to take better, worse, etc., as suppletive degrees of comparison to the corresponding adjectives.