I feel that I ought to mention a clear although more historically based interpretation which I haven't been able to consider here. This is the view of the musical idea which, instead of being seen as a teliological, dialectical process, is seen as an object which may be 'viewed' from different perspectives. In western music this has its roots in the music of Satie, and perhaps before that, in some of the nationalist styles. When combined with oriental musics and aesthetics this resulted in minimalism (phasing) and, in popular musics, ambience, etc. pSY can produce more or less interesting results in this format in at least one way (although I haven't done much investigation myself) as it 'liberates' the free editing of timbre while maintaining (a sort of) control over general MIDI functions.
By far the greatest proportion of music software is 'tool' based - it is designed to function as a tool for musicians to use rather than specifically providing a means of composition or other form of creative process. This is no place for a review of currently available software, (which in any case, is far too profuse to do anyway), but there are some general comments that can be made concerning general trends. As has been mentioned above, there are now programmes emerging which do 'create', although many of these tend to be what might be called 'experiments' or 'toys', and some larger software has creative functions included. It has become increasingly clear that as music software develops the boundary between 'tool' and 'creating device' is becoming increasingly blurred, and, moreover, that our use of a particular interface can have a significant effect on the output created by it. It is worth investigating this concept in slightly more detail and also, investigating to what extent it might be possible to diffuse or even eradicate this influence, (should it be seen as a bad thing?)
It could be argued that the software we use to create music is similar to a 'language' or at least one of a set of tools that are available to us. According to this argument, although the use of one particular language may have an effect on the manner in which things are said, and the tone, most (but far from all) people would suggest that the important part is what is said, or the concepts behind what is said rather than the manner in which it is said. This is not the place for a discussion of the nature of music itself - is it a language, a game, an abstraction, a cultural phenomenon, although without doubt our view of this will have an effect on how we use music in our lives.
Currently available music software is whether or not they are commercial products, or as is more and more relevant how commercially based they are. At one end are the 'fully functioning', quite highly priced and usually protected software products which can be equated in musical terms to the major word processors. They undertake the majority of functions normally considered necessary in the commercial market, and after having used one for a short time it usually becomes fairly apparent that this is the market at which they are aimed. Open one and almost certainly by default you will be presented with some sort of document representing your blank sheet of paper, except stencilled in will be a musical stave (or equivalent), no key signature (which is assumed to mean C major) and a 4/4 time signature. Depending on which package you are using these defaults will be more or less difficult to alter, but in general they will be more rather than less. However, again in general, if you want to perform a typically commercial activity, (quantisation for example) these will be quite straightforward. The reasons for these emphases is quite clear and uncontroversial - the majority of users of such software will be principally interested in creating commercial music and so, as these software companies depend on the sales of their software to such people, they would be foolish to risk their incomes in order to 'provoke' their clients into thinking a little more for themselves. This is analogous to the appending of (music) keyboards onto otherwise entirely unrelated sound boxes to make synthesisers as discussed above (The Musical Interface, p21). This all seems fairly harmless - I myself have written a number of acoustic pieces with the aid of sequencers such as Cubase and struggled (not always successfully) to get the programme to quantise in 'strange' non-commercial ways (hep- and septuplets are tricky), although I have also used cut and paste to create and develop ideas in ways that I don't think would have occurred to me without the use of the software. On the negative side, they enabled me to create interesting ideas and textures which turned out to be technically on or over the edge for live performers playing acoustic instruments. Similarly, there is a distinct move on the part of some composition students to rely almost entirely on a notator's abilities, often in an entirely unjustified way. There is a clear belief that if a piece of music is 'printed' it must be better than the same piece written by hand. There is every evidence that this view is both dangerous and untrue.
Although the boundary is blurred it is relatively clear that at the most commercial end, the reason for producing, developing and supporting this software is income based, although even here, there will be significant artistic interest. In these cases the very nature of artistic interest itself plays a part, for the artistic purpose of this software is, indeed, to enable people to express themselves, but in a commercially viable manner. This is, of course, a part of what most users want. How intermingled the ideas of being popular, communicating with a larger audience and making money actually are is a fascinating question and one without a single answer in the case of a single individual. It is, in any case, too large a subject to be covered here.
From one perspective, there seems nothing against such a laissez faire attitude to use of software. There is a natural tendency to believe that software developed for non-commercial, or at least non-profit-making purposes will be more pure of 'stylistic attitude', although I suppose that there is an argument opposing this use of the term style in the sense that it could be said that a style is inevitable in aesthetic terms and that in stating the above I am not taking this into account with regard to my own views. However, I would suggest that, as I mentioned above with regard to composing for acoustic instruments (p****), in my view an essential aspect of composing for any instrument involves taking full account of the characteristics of that instrument which in the case of most advanced synthesisers means accepting that the commonly supplied hardware interface (a music keyboard - often implemented in software in computer applications) is itself an anachronism that has nothing in reality to do with the processes of sound production inside the machine. In order to write for these hardware and software synthesisers fully, I would argue that our present forms of control, at least in performance terms are not just inadequate but fundamentally misleading. Many software packages to attempt to reinterpret the way we 'see' sound and some are more successful than others - at least in the non-commercial environment there is not so much pressure to produce something that is immediately and widely more successful as in the majority, as was mentioned above, the latter software is produced for the production of music that in general remains quite restricted in terms of the pitch and rhythms variety used.
However even in this field there has been no satisfactory agreement over 'ideal' ways of implementing sound visually. Admittedly, the area is really quite young and so, as with the development of 'norms' as in synthesiser design and implementation it is unlikely that there would be as yet, and the tendency is at the moment for many non-commercial (and commercial) developers to dash around trying to come up with the 'perfect' interface. There is the strongest hint at the moment that, as indicated by Penrose's comments on the variability in thought between different people (below, p35****), that this is a chimera.
Ultimately, the only way around the 'tool' with an attitude is to include programmable elements. There are elements of this already in most leading packages - the use of sliders, knobs, dial, etc., to which may be assigned MIDI messages. There are even some elements of number manipulating functions - for instance the Phrase Synthesiser in Cubase - and one could be fairly sure during the last ten or so years that a number of characteristically rhythmic passages in some commercial productions were using this or a related technology. In some audio packages, one can group together a number of functions into a 'script' or a 'macro', and this becomes, in a sense, a custom function.