9

This might seem like a very stupid question and it probably is, but hopefully not as much as one could expect.

My question in fact is the following: Does the difference between sum and product lie only in the distinction of their respective neutral elements (i.e. "0" for addition and "1" for multiplication) and in the "one-way" distributive property of multiplication over addition?

As a newbie coming to this subject my first instinct while reading up on the basics of abstract algebra (AA from here on) was to take literally the "abstract" in AA, which — as I take it — doesn't really have to do with numbers per se, but "the structures arising from different ways of combining things together", of which numbers are a very useful and recurring instance, but not the only one.

For instance i have learned (from a 3b1b video on youtube) that the set of isomorphic rotations of a polygon is a group, since it is closed under addition, where addition is computed by applying one rotation after the other (also commutativity applies, there exists a "zero" element, and for every non-zero element there exists the additive inverse of that element). Here "addition" seemed to me to have nothing to do with the familiar algebraic summation of numbers, except for the properties that it holds.

Is this the right way of interpreting the subject?
That is, am i right in believing that the definition of an operation (a mapping of a set upon itself of the kind $S\times S \longrightarrow S$) exactly coincides with its properties and nothing more? (setting aside the fact that we can elaborate different methods of computing the output of these binary operations on different objects)

ftv0
  • 111
  • I urge you to correct me at any point where in my writing I phrase something even in a slightly imprecise way since i am very new to abstract algebra and I might unconsciously fall into the paralogisms typical of a new learner. Also, this is my first question on the site and I haven't had much experience to know if this is the kind of question that is allowed to be on here or not. In the latter case I apologise profusely to the people who wasted time reading this and I promise I will look better into the functioning of the site before writing again. – ftv0 Sep 26 '23 at 19:23
  • 3
    There are no hard and fast rules which dictate how to denote binary operations, but there are some conventions which are widely used. One big one is that $+$ is reserved for commutative operations. That is, $a+b=b+a$ in whatever context the binary operation is appearing. Multiplication may be commutative (as standard multiplication or real numbers is) or it may not (as, for example, in the case of matrix multiplication). – lulu Sep 26 '23 at 19:27
  • 2
    @ftv0: FWIW I think this is a great question, this strikes me as a very natural thing to be confused about for a beginner to abstract algebra and there's really something important here worth explaining. Welcome! – Qiaochu Yuan Sep 26 '23 at 19:27
  • 3
    It is strange that 3b1b called this an addition. This is a nonabelian group under composition; it would be more normal to call this a group multiplication. But hey, that's purely stylistic. Remember that things like "$+$" and "$\cdot$" are placeholders for functions $G\times G\to G$ and then it technically doesn't matter. I could write my group law as $(a,b)\mapsto a-b$ - that would just be really weird and unconventional. Or as $a\pitchfork b$ or $a\obox b$ ... and so on. – FShrike Sep 26 '23 at 19:32
  • 1
    Yes, the distributive law is what distinguishes which operation is called addition vs. multiplication. In some structures they both may distribute over each other so there may be no such preference (e.g. Boolean algebras, distributive lattices). Normally addition is reserved for commutative operations (for rings: commutativity of addition is implied by distributivity if addition is cancellative). – Bill Dubuque Sep 26 '23 at 20:14
  • 2
    Fyi: accepting an answer quickly greatly reduces your chances of obtaining further answers (many users skip questions already having accepted answers). Better to wait at least a couple days so most everyone has a chance to see it. – Bill Dubuque Sep 26 '23 at 20:18

2 Answers2

10

This is subtle. You could say that the lesson of abstract algebra is that there is this more fundamental concept of an "operation," and both "addition" and "multiplication" are names we give to certain operations, first of all

  • literally addition and multiplication of numbers in the familiar sense, and second of all
  • operations that remind us of addition and multiplication in the familiar sense.

For example we also add functions, random variables, matrices, etc. and we also multiply matrices, elements of an arbitrary group, etc.

Conventionally it's common to think of the group operation of an abelian group as "addition" and the group operation of a not-necessarily-abelian group as "multiplication" but strictly speaking it's not necessary to do this and we do this only to guide our intuition. If you wanted to be very rigorous and careful you could insist on only talking in terms of "operations" here.

Also conventionally when we are talking about a ring we call one of its operations "addition" and the other one "multiplication" and multiplication is the one that distributes over addition. (This convention is compatible with the previous one; a ring has an underlying abelian group given by its "addition," as well as a multiplicative not-necessarily-abelian group given by its "multiplication" on invertible elements.) These operations generalize addition and multiplication of numbers but may look very different; it would be very annoying to have to come up with completely new names for them in general so we just keep calling them addition and multiplication. As you learn the subject you learn which intuitions about ordinary addition and multiplication carry over to this more abstract setting and which don't.

A major reason we don't make hard and fast distinctions between "additions" and "multiplications" in abstract algebra is that there are (sometimes) homomorphisms relating them, so they're not as distinct as they might appear. For example the exponential $x \mapsto \exp(x)$ is a group homomorphism, and in fact an isomorphism, from the real numbers $\mathbb{R}$ under addition to the positive reals $\mathbb{R}_{>0}$ under multiplication, with inverse the logarithm! This is a fancy way of talking about the familiar exponent and logarithm rules $\exp(a + b) = \exp(a) \exp(b)$ and $\log(ab) = \log(a) + \log(b)$.

Qiaochu Yuan
  • 419,620
0

Abstract algebra indeed refers to abstract entities, which need not be integers or other classes of numbers, though numbers and certain operations are particularly clear and useful examples. In group theory, for instance, abstract entities must obey a set of requirements (closure, existence of identity, inverses, associativity, ...) and can be represented fully by a "multiplication table" where 'multiplication' need not be the familiar operation on two numbers. For instance, the elements might correspond to configurations of a geometric figure and the operations to transformations such as clockwise rotation by $90^\circ$, flip about a vertical axis, and such.

Of course traditional addition and multiplication of numbers have different such structures for instance concerning inverses.

When I learned group theory, the professor generally used the symbol $\circ$ rather than $+$ or $*$ to be clear the operation in a group need not carry the meaning of traditional operations on numbers.

David G. Stork
  • 29,774