Cfrgtkky

I'm composing for orchestra.

I've read that divisions [of parts] should be notated using divisi. (or div.) above the staff. But I'm wondering whether this notation is mandatory for all instruments.

I can see why the notation is needed for strings to clear ambiguity against double-stopping but is there any rule regarding brass and woodwind instruments? Could the division be implied without the use of divisi notation?

(As far as I know, brass and woodwind instruments can't double-stop so divisions can't be ambiguous, but some affirmation would be nice.)

The usual situation, if there is more than one part for the same instrument in a piece, is that the instruments do not play the same note. Such parts can be, and usually are, notated on the same staff in an orchestral score. Each such staff can be extracted into a single part or into multiple parts, as described by MattPutnam. Having a separate part for each player is the norm, as noted by Michael Seifert in a comment.

Since it's common for a single staff to carry two parts, it is not customary to write divisi on the staff when there is more than one line of notes. Instead, the composer or copyist must indicate what to do in the opposite case, when there's only a single line of notes.

If the single line of notes is supposed to be played by more than one part, the usual marking is not unison, but a due or, for a staff carrying three parts, a tre. This is frequently written as a 2 or a 3. An example may be found in the German Requiem of Johannes Brahms, in the bassoons, in the second movement, at measure 1 (I'm looking at the Dover reprint of the Breitkopf and Härtel complete works).

If there is a single note that only one part should play, the marking is primo or secondo, often written as 1º, Iº, 1., or I. and 2º, IIº, 2., or II., using Italian or German conventions for writing ordinal numbers. An example may be found in the Brahms, in the first movement, at measure 45, in the trombones. Another possibility is to write a rest under the note, if the note is for the first player, or above it, if the note is for the second player. An example of this is seen in the second movement at measures 33 and 34, in the oboes and bassoons.

If the parts combine for just a few notes, the convention is to write each note with both upward and downward pointing stems. This can be seen in the second movement of the Brahms, in the bassoons, from measure 13 through measure 22. For notes that have no stem, the convention is to write two notes side by side. This can be seen in the last measure of the same movement, on the third-trombone-and-tuba staff.

Finally, to underscore the fact that there is no need to mark divisi explicitly, look again at the beginning of the same movement, where the bassoons are a due. On the third beat of measure 12, they divide without comment, and on the downbeat of the next measure, where they are again in unison, the note has two stems. The parts remain divided until measure 42, where the staff is marked a 2.

I will close with an endorsement of Carl Witthoft's answer, which applies in the context of writing for wind and brass sections as is perhaps more common for bands. The assumption in orchestral writing is that each woodwind and brass part is assigned to a single player. There are certainly cases in which orchestral wind and brass parts have been doubled, such as at the premiere of Beethoven's ninth symphony. In such a case it would be for the conductor to decide whether any passage should be played by the entire section or as a solo. A composer could explicitly call for multiple-player sections in an orchestral piece, and then markings like solo, tutti, divisi, and unison would be useful, but that is not the convention.

First, be aware that there's often a difference between the conductor's score and the individual parts. To save space and make it easier to navigate, a score often has multiple parts condensed onto one staff, even if the result would be horribly complex to decipher while playing.

The individual parts can be written either combined or separate, depending on what makes more sense. If the parts generally move in the same rhythm and don't cross, then it's not a problem to combine them. Or if there's tricky interplay between the parts, it may be beneficial to combine them on the page. But if the parts are very independent, then they should be separated.

So the roundabout answer to your question is that the parts should be written in a way that such markings aren't needed. Obviously, a wind instrument can only play one note at a time (in general, yes I know about multiphonics), so if there are two notes they will divide. But if you are trying to write a combined part and find yourself needing to write a bunch of unison/1st only/2nd only instructions, then that's a sign that you should be writing two separate parts.

It never hurts to be clear even if mildly redundant. Consider one possibility: you write two parts for a wind section and mark (intentionally) the upper line as "solo," to be played by one person with the remainder of the section playing the other line (this makes more sense in a wind ensemble/concert band than an orchestra). The solo ends; everyone plays the single line; then later on you want a divisi. If you don't indicate, they might think it's another solo line.