FUGUE (Lat. fuga, flight), in music, the mutual "pursuit" of voices or parts. It was, up to the end of the 16th century, if not later, the name applied to two art-forms. (A) Fuga ligata was the exact reproduction by one or more voices of the statement of a leading part. The reproducing voice (comes) was seldom if ever written out, for all differences between it and the dux were rigidly systematic; e.g. it was an exact inversion, or exactly twice as slow, or to be sung backwards, etc. etc. Hence, a rule or canon was given, often in enigmatic form, by which the comes was deduced from the dux: and so the term canon became the appropriate name for the form itself, and is still retained. (B) A composition in which the canonic style was cultivated without canonic restriction was, in the 16th century, called fuga ricercata or simply a ricercare, a term which is still used by Bach as a title for the fugues in Das musikalische Opfer.
The whole conception of fugue, rightly understood, is one of the most important in music, and the reasons why some contrapuntal compositions are called fugues, while others are not, are so trivial, technically as well as aesthetically, that we have preferred to treat the subject separately under the general heading of Contrapuntal Forms, reserving only technical terms for definition here.
(i.) If in the beginning or "exposition" the material with which the opening voice accompanies the answer is faithfully reproduced as the accompaniment to subsequent entries of the subject, it is called a countersubject (see Counterpoint, under sub-heading Double Counterpoint). Obviously the process may be carried further, the first countersubject going on to a second when the subject enters in the third part and so on. The term is also applied to new subjects appearing later in the fugue in combination (immediate or destined) with the original subject. Cherubini, holding the doctrine that a fugue cannot have more than one subject, insists on applying the term to the less prominent of the subjects of what are commonly called double fugues, i.e. fugues which begin with two parts and two subjects simultaneously, and so also with triple and quadruple fugues.
(ii.) Episodes are passages separating the entries of the subject.  Episodes are usually developed from the material of the subject and countersubjects; they are very rarely independent, but then conspicuously so.
(iii.) Stretto, the overlapping of subject and answer, is a resource the possibilities of which may be exemplified by the setting of the words omnes generationes in Bach's Magnificat (see Bach).
(iv.) The distinction between real and tonal fugue, which is still sometimes treated as a thing of great historical and technical importance, is really a mere detail resulting from the fact that a violent oscillation between the keys of tonic and dominant is no part of the function of a fugal exposition, so that the answer is (especially in its first notes and in points that tend to shift the key) not so much a transposition of the subject to the key of the dominant as an adaptation of it from the tonic part to the dominant part of the scale, or vice versa; in short, the answer is as far as possible on the dominant, not in the dominant. The modifications this principle produces in the answer (which have been happily described as resembling "fore-shortening") are the only distinctive marks of tonal fugue; and the text-books are half filled with the attempt to reduce them from matters of ear to rules of thumb, which rules, however, have the merit (unusual in those of the academic fugue) of being founded on observation of the practice of great masters. But the same principle as often as not produces answers that are exact transpositions of the subject; and so the only kind of real fugue (i.e. fugue with an exact answer) that could rightly be contrasted with tonal fugue would be that in which the answer ought to be tonal but is not. It must be admitted that tonal answers are rare in the modal music of the 16th century, though their melodic principles are of yet earlier date; still, though tonal fugue does not become usual until well on in the 17th century, the idea that it is a separate species is manifestly absurd, unless the term simply means "fugue in modern tonality or key," whatever the answer may be.
The term "answer" is usually reserved for those entries of the subject that are placed in what may be called the "complementary" position of the scale, whether they are "tonally" modified or not. Thus the order of entries in the exposition of the first fugue of the Wohltemp. Klav. is subject, answer, answer, subject; a departure from the usual rule according to which subject and answer are strictly alternate in the exposition.
In conclusion we may remind the reader of the most accurate as well as the most vivid description ever given of the essentials of a fugue, in the famous lines in Paradise Lost, book xi.
"His volant touch,
Instinct through all proportions, low and high,
Fled and pursued transverse the resonant fugue."
It is hard to realize that this description of organ-music was written in no classical period of instrumental polyphony, but just half-way between the death of Frescobaldi and the birth of Bach. Every word is a definition, both retrospective and prophetic; and in "transverse" we see all that Sir Frederick Gore Ouseley expresses in his popular distinction between the "perpendicular" or homophonic style in which harmony is built up in chords, and the "horizontal" or polyphonic style in which it is woven in threads of independent melody.
(D. F. T.)
 An episode occurring during the exposition is sometimes called codetta, a distinction the uselessness of which at once appears on an analysis of Bach's 2nd fugue in the Wohltemp. Klav. (the term codetta is more correctly applied to notes filling in a gap between subject and its first answer, but such a gap is rare in good examples).
Note - this article incorporates content from Encyclopaedia Britannica, Eleventh Edition, (1910-1911)