[G.Polya]
Terms, old and new, Describing the activity of solving problems are often ambiguous. The activity itself is familiar to everybody and it is often discussed but, as other mental activities, it is difficult to describe. In the absence of a systematic study there are no technical terms to describe it, and certain usual half technical terms often add to the confusion because they are used in different meanings by different authors. The following short list includes a few new terms used and a few old terms avoided in the present study, and also some old terms retained despite their ambiguity. The reader may be confused by the following discussion of terminology unless his notions are well anchored in examples. 1. analysis is neatly defined by PAPPUS and it is a useful term describing a typical way of devising a plan, starting from the unknown (or the conclusion) and working backwards, toward the data (or the hypothesis). Unfortunately the word has acquired very different meanings (for instance, of mathematical, chemical, logical analysis) and therefor it is regretfully avoided in the present study. 2. Condition links the unknown of a "problem to find" to the data (see PROBLEMS TO FIND, PROBLEMS TO PROVE,3) In this meaning it is clear, useful and unavoidable term. It is often necessary to decompose the condition into several parts [into parts (I) and (II) in the examples DECOMPOSING AND RECOMBINING?,7,]. Now, this ambiguity which is sometimes embarrassing could be easily avoided by introducing some technical term to denote the parts of the whole condition; for instance, such a part could be called a "clause." 3. Hypothesis denotes an essential part of a mathematical theorem of the more usual kind(see PROBLEMS TO FIND, PROBLEMS TO PROVE, 4). The term, in this meaning, is perfectly clear and satisfactory. The difficulty is that each part of the Hypothesis is also called a hypothesis so that the hypopthesis mat consist of several hypotheses The remedy would be to call each part of the whole hypothesis a "clause," or something similar. (Compare the foregoing remark on "condition.") 4. Principal partsPROBLEMS TO FIND, PROBLEMS TO PROVE,3,4.) 5. Problem to find, problem to prove are a pair of new terms, introduced regretfully to replace historical terms whose meaning, however, is confused beyond redemption by current usage. In Latin versions of Greek mathematical texts, the common name for both kinds of problems is "proposition" a "problem to find" is called "problem" and a "problem to prove" "theorem" In old-fashioned mathematical language, the words proposition, problem theorem have still this "Euclidean" meaning, gut this is completely changed in modern mathematical language; this justifies the introduction of new terms. 6. Progressive reasoning was used in various meanings by various authors, and in the old meaning of "synthesis" (see 9)by some authors. The latter usage is defensible but the terms avoided here. 7. Regressive reasoning was used in the old meaning of "analysis" by some authors (compare 1,6). The term is defensible but avoided here. 8. Solution is a completely clear term if taken in its purely mathematical meaning; it denotes any object satis- fying the condition of a "problem to find." Thus, the solutions of the equation x^2-3x+2=0 are its roots, the numbers 1 and 2. Unfortunately, the word has also other meanings which are not purely mathematical and which are used by mathematical along with its mathematical meaning. Solution may also mean the "process of solving the problem or the "work done in solving the problem"; we use the word in this meaning when we talk about a "difficult solution." Solution may also mean the result of the world done in solving the problem; we may use the word in this meaning when we talk about a "beautiful solution. Now, it may happen that we have to talk in the same sentence about the object satisfying the condition of the problem, about the work obtaining it and about the result of this work; if we yield to the temptation to call all three things "solution" the sentence cannot be too clear. 9. Synthesis is used by PAPPUS in a well defined meaning which would deserve to be conserved, The term is however, regretfully avoided in the present study, for the same reasons as its counterpart "analysis" (see under 1).