It is almost impossible for all nine types of sources to survive for a single work. For many works, one or more sources may be lost; for others, certain categories may never have existed in the first place. In any good modern edition, the editor should list all sources known to have existed, including any that may be missing. They should also make sure to consult all those available, either originals themselves or at least photocopies or microfilms of them.
Whenever a work from the Classical period survives in more than one source, there will almost certainly be difference in the text, known as “variants”. The editor should try to sort out the differences between them. A family tree called ‘stemmata’ shows these relationships. All editors working on text from Middle Ages to Classical Era would be aware of this problem for many decades and the study of the stemmata has become quite a sophisticated science. Music editing is mainly done by excellent performers who have no training in the scientific methods. Therefore, they merely consult vague manuscripts and original editions as it is rare for a modern edition to contain a full assessment of relationship between all the sources.
Depending on the works, some stemma will be quite straight forward while others can be quite complex especially if the composer has made a significant revision. It is rarely possible to verify that one source derived from the preceding sources unless there are some external evidence.
Even though variants are a great help in establishing the relationship between sources, it also presents an enormous problem when the editor is trying to re-create the original text. It requires a high degree of skill, knowledge, and musical judgement. Relying on only one of the several sources is almost as bad as relying random mixture of variants from different sources. The editor should work out the cause of each variant and thereby create his edition based on what he or she believes to be the composer’s final intention.
There are a number of possible variants which can be categorized as follows:
- Corruption of the text, caused by miscopying
- Degeneration of the text, caused by accidental omission of details such a slurs or ornaments
- Distortion, through a copyist’s deliberate alteration, perhaps through simplification or updating the music
- Correction, or attempted correction of obvious errors
- Revision made by composer
- Amplification of the text through additional performance indication such as metronome marks
- Faulty memorization of a work when it is being written out from memory
- Notational changes intended to clarity the appearance of the music without affecting its sound
It is not always possible to be sure which one is responsible for which variant. However, if an editor applies a systematic approach it can greatly assist the editor to derive the most reliable text. It is not always possible to determine which version was intended by the composer, so when the variant that was not incorporated into the editor’s final music text it should be listed as a critical commentary. Performers can therefore see how the text they are to perform were derived and form a different opinion.
Critical commentaries are generally dense and filled with abbreviations, making it daunting to read. Performers are encouraged to consult them especially where the musical text presented seem slightly odd.
Where the edition is based on only one or on several sources, there may be places where the sources contain an error. Some errors are more defined than others. A good editor will correct everything that is believed to be erroneous and indicate that this has been corrected. If amendment is more complicated, it should be mentioned in the critical commentary or introduction or by an explanatory note on the page itself.
The quality of the modern edition cannot be judge by the criteria whether the fingering is included, or the editor is a well-known performer nor can it be judged by the length of the commentary at the beginning or the end of the volume. Sometimes, the commentary is published separately as it isn’t practical to include dull critical commentary in an anthology of pieces reprinted from a variety of different modern editions.
There needs to be a brief reference to where the editorial material can be found. The date of a good edition can be found dating before 1950. As well, certain publishers and editors are renowned for producing editions of high quality. Nevertheless, each edition needs to be judged using the criteria and principles outlined above.