Category Archives: understanding math

Guest post by F. Kafi

Before I posted Extensional and Intensional, I had emailed a draft to F. Kafi.  The following was his response.  –cw

 

In your example, “Suppose you set out to prove that if $f(x)$ is a differentiable function and $f(a)=0$ and the graph going from left to right goes UP to $f(a)$ and then DOWN after that then $a$ has to be a maximum of the function”, could we have the graph of the function $f(x)$ without being aware of the internal structure of the function; i.e., the mathematical formulation of $f(x)$ such as $f(x):=-(x-a)^2$ or simply its intensional meaning? Certainly not.

Furthermore, what paves the way for the comparison with our real world experiences leading to the metaphoric thinking is nothing but the graph of the function. Therefore, it is the intensional meaning of the function which makes the metaphoric mode of thinking possible.

The intensional meaning is specially required if we are using a grounding metaphor. A grounding metaphor uses concepts from our physical and real world life. As a result we require a medium to connect such real life concepts like “going up” and “going down” to mathematical concepts like the function $f(x)$. The intensional meaning of function $f(x)$ through providing numbers opens the door of the mind to the outer world. This is possible because numbers themselves are the result of a kind of abstraction process which the famous educational psychologist Piaget calls empirical abstraction. In fact, through empirical abstraction we transform the real world experience to numbers.

 

 

Let’s consider an example. We see some racing cars in the picture above, a real world experience if you are the spectator of a car match. The empirical abstraction works something like this:

 

 

Now we may choose a symbol like "$5$" to denote our understanding of "|||||".

It is now clear that the metaphoric mode of thinking is the reverse process of “empirical abstraction”. For example, in comparing “|||||||||||” with “||||” we may say “A car race with more competing cars is much more exciting than a much less crowded one.” Therefore, “|||||||||||”>“||”, where “>” is the abstraction of “much more exciting than”.

In the rigorous mode of thinking, the idea is almost similar. However, there is an important difference. Here again we have a metaphor. But this time, the two concepts are mathematical. There is no outer world concept. For example, we want to prove a differentiable function is also a continuous one. Both concepts of “differentiability” and “continuity” have rigorous mathematical definitions. Actually, we want to show that differentiability is similar to continuity, a linking metaphor. As a result, we again require a medium to connect the two mathematical concepts. This time there is no need to open the door of the mind to the outer world because the two concepts are in the mind. Hence, the intensional meaning of function $f(x)$ through providing numbers is not helpful. However, we need the intensional meanings of differentiability and continuity of $f(x)$; i.e., the logical definitions of differentiability and continuity.

In the case of comparing the graph of $f(x$) with a real hill we associated dots on the graph with the path on the hill. Right? Here we need to do the the same. We need to associate the $f(x)$’s in the definition of differentailblity to the $f(x)$’s used in the definition of continuity. The $f(x)$’s play the role of dots on the graph. As the internal structure of dots on the graph are unimportant to the association process in the grounding metaphor, the internal structure of $f(x)$’s in the logical definition are unimportant to the association process in the linking metaphor. Therefore, we only need the extensional meaning of the function $f(x)$; i.e., syntactically valid roles it can play in expressions.

Extensional and Intensional

This post uses the word intensional, which is not the word "intentional" and doesn't mean the same thing.

 

The connection between rich view/rigorous view and intensional/extensional

 

In the abmath article Images and Metaphors I wrote about the rigorous view of math, in contrast to the rich view which allows metaphors, images and intuition. F. Kafi has proposed the following thesis:

The rigorous mode of thinking deals with the extensional meaning of mathematical objects while the metaphoric mode of thinking deals with the intensional meaning of mathematical objects.

This statement is certainly suggestive as an analogy. I have several confused and disjointed thoughts about it.

What does "intensional" mean?

Philosophy

Philosophers say that "the third largest planet in the solar system" has intensional meaning and "Neptune" has extensional meaning. Among other things we might discover a planet ridiculously far out that is bigger than Neptune. But the word "Neptune" denotes a specific object.

The intensional meaning of "the third largest planet in the solar system" has a hidden time dimension that, if made overt, makes the statement more nearly explicit. (Don't read this paragraph as a mathematical statement; it is merely thrashing about to inch towards understanding.)

Computing science

Computer languages are distinguishes as intensional or extensional, but their meaning there is technical, although clearly related to the philosophers' meaning.

I don't understand it very well, but in Type Theory and in Logic, an intensional language seems to make a distinction between declaring two math objects to be equal and proving that they are equal. In an extensional language there is no such distinction, with the effect that in a typed language typing would be undecidable.

Here is another point: If you define the natural numbers by the Peano axioms, you can define addition and then prove that addition is commutative. But for example a vector space is usually defined by axioms and one of the axioms is a declaration that addition of vectors is commutative. That is an imposed truth, not a deduced one. So is the difference between intensional and extensional languages really a big deal or just a minor observation?

What is "dry-bones rigor"?

Another problem is that I have never spelled out in more than a little detail what I mean by rigor, dry-bones rigor as I have called it. This is about the process mathematicians go through to prove a theorem, and I don't believe that process can be given a completely mathematical description. But I could go into much more detail than I have in the past.

Suppose you set out to prove that if $f(x)$ is a differentiable function and $f(a)=0$ and the graph going from left to right goes UP before $x$ reaches $a$ and then DOWN for $x$ to the right of $a$, then $a$ has to be a maximum of the function. That is a metaphorical description based on the solid physical experience of walking up to the top of a hill. But when you get into the proof you start using lots of epsilons and deltas. This abandons ideas of moving up and down and left to right and so on. As one of the members of Bourbaki said, rigorous math is when everything goes dead. That sounds like extensionality, but isn't their work really based on the idea that everything has to be reduced to sets and logic? (This paragraph was modified on 2013.11.07)

Many perfectly rigorous proofs are based on reasoning in category theory. You can define an Abelian group as a categorical diagram with the property that any product preserving functor to any category will result in a group. This takes you away from sets altogether, and is a good illustration of the axiomatic method. It is done by using nodes, arrows and diagrams. The group is an object and the binary operation is an arrow from the square of the object. Commutativity is required by stating that a certain diagram must commute. But when you prove that two elements in an Abelian group (an Abelian topological group, an Abelian group in the category of differentiable manifolds, or whatever) can be added in either order, then you find yourself staring at dead arrows and diagrams rather than dead collections of things and so you are still in rigor mortis mode.

I will write a separate post describing these examples in much more detail than you might want to think about.

Metaphors and intensionality

One other thing I won't go into now: How are thinking in metaphors and intensional descriptions related? It seems to me the two ideas are related somehow, but I don't know how to formulate it.

“Trivial”

This is an update of an article in the abstractmath glossary.

About theorems

A theorem is said to be trivial to prove or trivially true

Example:

Here is a scenario that exemplifies (a):

  • A textbook defines the image of a function $F:A\to B$ to be the set of all elements of $B$ of the form $F(a)$ for some $a\in A$.
  • It then goes on to say that $F$ is surjective if for every element $b$ of $B$ there is an element $a\in A$ with the property that $F(a) = b$.
  • It then states a theorem, or give an exercise, that says that a function $F:A\to B$ is surjective if and only if the image of $F$ is $B$.
  • The proof follows immediately by rewriting using definitions.
  • The instructor calls the proof trivial and goes on to the next topic.
  • Some students are totally baffled.

I have seen this happen many times with this and other theorems.  This sort of incident may be why many intelligent people feel they are "bad at math".

People are not born knowing the principle of rewriting by definitions. The principle needs to be TAUGHT.

  • When a class is first introduced to proof techniques the instructor should explicitly describe rewriting by definitions with several examples.
  • After that, the instructor can say that a proof follows by rewriting by definitions and make it clear that the students will have to do the work (then or later).
  • Such a proof is justly called "trival" but saying it is trivial is also a putdown if no one has pointed out the procedure of rewriting by definitions.

Example:

This example illustrates (b).

Theorem: Let $G$ be a finite group and $H$ a subgroup of index $2$ (meaning it has half the number of elements of the group).  Then $H$ is normal in $G$.

Basic facts about groups and subgroups learned in first semester abstract algebra:

  • A subgroup of a group determines a partition consisting of left cosets and another partition of right cosets, each (in the finite case) with the same number of elements as the subgroup.
  • A subgroup is a left coset of itself and also a right coset of itself.
  • If every left coset is also a right coset and vice versa (so the two partitions just mentioned are the same), then by definition the subgroup is normal in the group.

Now if $H$ has index $2$ that means that each partition consists of two cosets. In both cases, one of them has to be $H$, so the other one has to be $G\setminus H$, which must therefore be a left and right coset of $H$. So $H$ is normal in $G$.

So once you understand the basics about cosets and normal subgroups, the fact that $H$ has to be normal if it is of index $2$ is "obvious". I don't think you should call this "trivial". Best to say it is "obvious if you have a clear understanding of cosets of groups".

About mathematical objects

  • A function may be called trivial if it is the identity function or a constant function, and possibly in other circumstances. (If someone showed that the cosmo­logical constant is 0 that would not be called trivial.)
  • A solution to an equation may be said to be trivial if it is 0 or 1. There may be other situations in which a solution is called "trivial" as well.
  • A mathematical structure is said to be trivial if its underlying set is empty or a singleton set. In particular, a subset of a set is nontrivial if it is nonempty. I have not found an example where "nontrivial subset" means it is not a singleton. 

Note: "Trivial" and "degenerate" overlap in meaning but are not interchangeable.  What is called "degenerate" seems to depend on the mathematical specialty.

Thinking about a function as a mathematical object

A mathematician’s mental representation of a function is generally quite rich and may involve many different metaphors and images kept in mind simultaneously. The abmath article on metaphors and images for functions discusses many of these representations, although the article is incomplete. This post is a fairly thorough rewrite of the discussion in that article of the representation of the concept of “function” as a mathematical object. You must think of functions as math objects when you are taking the rigorous view, which happens when you are trying to prove something about functions (or large classes of functions) in general.

What often happens is that you visualize one of your functions in many of the ways described in this article (it is a calculation, it maps one space to another, its graph is bounded, and so on) but those images can mislead you. So when you are completely stuck, you go back to thinking of the function as an axiomatically-defined mathe­matical structure of some sort that just sits there, like a complicated machine where you can see all the parts and how they relate to each other. That enables you to prove things by strict logical deduction. (Mathematicians mostly only go this far when they are desperate. We would much rather quote somebody’s theorem.) This is what I have called the dry bones approach.

The “mathematical structure” is most commonly a definition of function in terms of sets and axioms. The abmath article Specification and definition of “function” discusses the usual definitions of “function” in detail.

Example

This example is intended to raise your consciousness about the possibilities for functions as objects.

Consider the function $f:\mathbb{R}\to\mathbb{R}$ defined by $f(x)=2{{\sin }^{2}}x-1$. Its value can be computed at many different numbers but it is a single, static math object.

You can apply operators to it

  • Just as you can multiply a number by $2$, you can multiply $f$ by $2$.   You can say “Let $g(x)=2f(x)$” or “Let $g=2f$”. Multiplying a numerical function by $2$ is an operator that take the function $f$ to $2f$. Its input is a function and its output is another function. Then the value of $g$ (which is $2f$) at any real $x$ is $g(x)=2f(x)=4{{\sin }^{2}}x-2$. The notation  “$g=2f$” reveals that mathematicians think of $f$ as a single math object just as the $3$ in the expression “$2\times 3$” represents the number $3$ as a single object.
  • But you can’t do arithmetic operations to functions that don’t have numerical output, such as the function $\text{FL}$ that takes an English word to its first letter, so $\text{FL}(`\text{wolf’})=`\text{w’}$. (The quotes mean that I am writing about the word ‘wolf’ and the letter ‘w’.) The expression $2\times \text{FL}(`\text{wolf’})$ doesn’t make sense because ‘w’ is a letter, not a number.
  • You can find the derivative.  The derivative operator is a function from differentiable functions to functions. Such a thing is usually called an operator.  The derivative operator is sometimes written as $D$, so $Df$ is the function defined by: “$(Df)(x)$ is the slope of the tangent line to $f$ at the point $(x,f(x)$.” That is a perfectly good definition. In calculus class you learn formulas that allow you to calculate $(Df)(x)$ (usually called “$f’(x)$”) to be $4 \sin (x) \cos (x)$.

Like all math objects, functions may have properties

  • The function defined by $f(x)=2{{\sin}^{2}}x-1$ is differentiable, as noted above. It is also continuous.
  • But $f$ is not injective. This means that two different inputs can give the same output. For example,$f(\frac{\pi}{3})=f(\frac{4\pi}{3})=\frac{1}{2}$. This is a property of the whole function, not individual values. It makes no sense to say that $f(\frac{\pi}{3})$ is injective.
  • The function $f$ is periodic with period $2\pi$, meaning that for any $x$, $f(x+2\pi)=f(x)$.     It is the function itself that has period $2\pi$, not any particular value of it.  

As a math object, a function can be an element of a set

  • For example,$f$ is an element of the set ${{C}^{\infty }}(\mathbb{R})$ of real-valued functions that have derivatives of all orders.
  • On ${{C}^{\infty }}(\mathbb{R})$, differentiation is an operator that takes a function in that set to another function in the set.   It takes $f(x)$ to the function $4\sin x\cos x$.
  • If you restrict $f$ to the unit interval, it is an element of the function space ${{\text{L}}^{2}}[0,1]$.   As such it is convenient to think of it as a point in the space (the whole function is the point, not just values of it).    In this particular space, you can think of the points as vectors in an uncountably-infinite-dimensional space. (Ideas like that weird some people out. Do not worry if you are one of them. If you keep on doing math, function spaces will seem ordinary. They are OK by me, except that I think they come in entirely too many different kinds which I can never keep straight.) As a vector, $f$ has a norm, which you can think of as its length. The norm of $f$ is about $0.81$.

The discussion above shows many examples of thinking of a function as an object. You are thinking about it as an undivided whole, as a chunk, just as you think of the number $3$ (or $\pi$) as just a thing. You think the same way about your bicycle as a whole when you say, “I’ll ride my bike to the library”. But if the transmission jams, then you have to put it down on the grass and observe its individual pieces and their relation to each other (the chain came off a gear or whatever), in much the same way as noticing that the function $g(x)=x^3$ goes through the origin and looks kind of flat there, but at $(2,8)$ it is really rather steep. Phrases like “steep” and “goes through the origin” are a clue that you are thinking of the function as a curve that goes left to right and levels off in one place and goes up fast in another — you are thinking in a dynamic, not a static way like the dry bones of a math object.

The definition of “function”

 

This is the new version of the abstractmath article on the definition of function. I had to adapt the formatting and some of it looks weird, but legible. It is prettier on abstractmath.org.

I expect to announce new revisions of other abmath articles on this blog, with links, but not to publish them here. This article brings out a new point of view about defining functions that I wanted to call attention to, so I am publishing it here, as well.

 

FUNCTIONS: SPECIFICATION AND DEFINITION

It is essential that you understand many of the images, metaphors and terminology that mathe­maticians use when they think and talk about functions. For many purposes, the precise mathematical definition of "function" does not play much of a role when you are trying to understand particular kinds of functions. But there is one point of view about functions that has resulted in fundamental progress in math:

 

 

A function is a mathematical object.

To deal with functions in that way you need a precise definition of "function". That is what this article gives you.

  • The article starts by giving a specification of "function".
  • After that, we get into the technicalities of the definitions of the general concept of function.
  • Things get complicated because there are several inequivalent definitions of "function" in common use.

Specification of "function"

A function $f$ is a mathematical object which determines and is completely determined by the following data:

(DOM) $f$ has a domain, which is a set. The domain may be denoted by $\text{dom} f$.

(COD) $f$ has a codomain, which is also a set and may be denoted by $\text{cod} f$.

(VAL) For each element $a$ of the domain of $f$, $f$ has a value at $a$, denoted by $f(a)$.

(FP) The value of $f$ at $a$ is completely determined by $a$ and $f$.

(VIC) The value of $f$ at $a$ must be an element of the codomain of $f$.

  • The operation of finding $f(a)$ given $f$ and $a$ is called evaluation.
  • "FP" means functional property.
  • "VIC" means "value in codomain".

Examples

The examples of functions chapter contains many examples. The two I give here provide immediate examples.

A finite function

Let $F$ be the function defined on the set $\left\{1,\,2,3,6 \right\}$ as follows: $F(1)=3,\,\,\,F(2)=3,\,\,\,F(3)=2,\,\,\,F(6)=1$. This is the function called "Finite'' in the chapter on examples of functions.

  • The definition of $F$ says "$F$ is defined on the set $\left\{1,\,2,\,3,\,6 \right\}$". That phrase means that the domain is that set.
  • The value of $F$ at each element of the domain is given explicitly. The value at 3, for example, is 2, because the definition says that $F(2) = 3$. No other reason needs to be given. Mathematical definitions can be arbitrary.
  • The codomain of $F$ is not specified, but must include the set $\{1,2,3\}$. The codomain of a function is often not specified when it is not important — which is most of the time in freshman calculus (for example).

A real-valued function

Let $G$ be the real-valued function defined by the formula $G(x)={{x}^{2}}+2x+5$.

  • The definition of $G$ gives the value at each element of the domain by a formula. The value at $3$, for example, is $G(3)=3^2+2\cdot3+5=20$.
  • The definition of $G$ does not specify the domain. The convention in the case of functions defined on the real numbers by a formula is to take the domain to be all real numbers at which the formula is defined. In this case, that is every real number, so the domain is $\mathbb{R}$.
  • The definition does not specify the codomain, either. However, must include all real numbers greater than or equal to 4. (Why?)

What the specification means

  • The specification guarantees that a function satisfies all five of the properties listed.
  • The specification does not define a mathematical structure in the way mathematical structures have been defined in the past: In particular, it does not require a function to be one or more sets with structure.
  • Even so, it is useful to have the specification, because:

     

     

    Many mathematical definitions
    introduce extraneous technical elements
    which clutter up your thinking
    about the object they define.

     

     

    I will say more about this when I give the various definitions that are in use.

History

Until late in the nineteenth century, functions were usually thought of as defined by formulas (including infinite series). Problems arose in the theory of harmonic analysis which made mathematicians require a more general notion of function. They came up with the concept of function as a set of ordered pairs with the functional property (discussed below), and that understanding revolutionized our understanding of math.

This discussion is an over­simpli­fication of the history of mathe­matics, which many people have written thick books about. A book relevant to these ideas is Plato's Ghost, by Jeremy Gray.

In particular, this definition, along with the use of set theory, enabled abstract math (ahem) to become a common tool for understanding math and proving theorems. It is conceivable that some of you may wish it hadn't. Well, tough.

The more modern definition of function given here (which builds on the older definition) came into use beginning in the 1950's. The strict version became necessary in algebraic topology and is widely used in many fields today.

The concept of function as a formula never disappeared entirely, but was studied mostly by logicians who generalized it to the study of function-as-algorithm. Of course, the study of algorithms is one of the central topics of modern computing science, so the notion of function-as-formula (updated to function-as-algorithm) has achieved a new importance in recent years.

To state both the old abstract definition and the modern one, we need a preliminary idea.

The functional property

A set $P$ of ordered pairs has the functional property if two pairs in $P$ with the same first coordinate have to have the same second coordinate (which means they are the same pair). In other words, if $(x,a)$ and $(x,b)$ are both in $P$, then $a=b$.

How to think about the functional property

The point of the functional property is that for any pair in the set of ordered pairs, the first coordinate determines what the second one is. That's why you can write "$G(x)$'' for any $x $ in the domain of $G$ and not be ambiguous.

Examples

  • The set $\{(1,2), (2,4), (3,2), (5,8)\}$ has the functional property, since no two different pairs have the same first coordinate. Note that there are two different pairs with the same second coordinate. This is irrelevant to the functional property.
  • The set $\{(1,2), (2,4), (3,2), (2,8)\}$ does not have the functional property. There are two different pairs with first coordinate 2.
  • The empty set $\emptyset$ has the function property vacuously.

Graph of a function.

Example: graph of a function defined by a formula

In calculus books, a picture like this one (of part of $y=x^2+2x+5$) is called a graph. Here I use the word "graph" to denote the set of ordered pairs \[\left\{ (x,{{x}^{2}}+2x+5)\,\mathsf{|}\,x\in \mathbb{R } \right\}\] which is a mathematical object rather than some ink on a page or pixels on a screen.

The graph of any function studied in beginning calculus has the functional property. For example, the set of ordered pairs above has the functional property because if $x$ is any real number, the formula ${{x}^{2}}+2x+5$ defines a specific real number.

  • if $x = 0$, then ${{x}^{2}}+2x+5=5$, so the pair $(0, 5)$ is an element of the graph of $G$. Each time you plug in $0$ in the formula you get 5.
  • if $x = 1$, then ${{x}^{2}}+2x+5=8$.
  • if $x = -2$, then ${{x}^{2}}+2x+5=5$.

You can measure where the point $\{-2,5\}$ is on the (picture of) the graph and see that it is on the blue curve as it should be. No other pair whose first coordinate is $-2$ is in the graph of $G$, only $(-2, 5)$. That is because when you plug $-2$ into the formula ${{x}^{2}}+2x+5$, you get $5$ and nothing else. Of course, $(0, 5)$ is in the graph, but that does not contradict the functional property. $(0, 5)$ and $(-2, 5)$ have the same second coordinate, but that is OK.

Modern mathematical definition of function

A function $f$ is a mathematical structure consisting of the following objects:

  • A set called the domain of $f$, denoted by $\text{dom} f$.
  • A set called the codomain of $f$, denoted by $\text{cod} f$.
  • A set of ordered pairs called the graph of $ f$, with the following properties:
  • $\text{dom} f$ is the set of all first coordinates of pairs in the graph of $f$.
  • Every second coordinate of a pair in the graph of $f$ is in $\text{cod} f$ (but $\text{cod} f$ may contain other elements).
  • The graph of $f$ has the functional property.

Using arrow notation, this implies that $f:A\to B$.

Remark

The main difference between the specification of function given previously and this definition is that the definition replaces the statement "$f$ has a value at $a$" by introducing a set of ordered pairs (the graph) with the functional property.

  • This set of ordered pairs is extra structure introduced by the definition mainly in order to make the definition a classical sets-with-structure, which makes the graph, which should be a concept derived from the concept of function, into an apparently necessary part of the function.
  • That suggests incorrectly that the graph is more of a primary intuition that other intuitions such as function as relocator, function as transformer, and other points of view discussed in the article Intuitions and metaphors for functions.

Examples

  • Let $F$ have graph $\{(1,2), (2,4), (3,2), (5,8)\}$ and define $A = \{1, 2, 3, 5\}$ and $B = \{2, 4, 8\}$. Then $F:A\to B$ is a function. In speaking, we would usually say, "$F$ is a function from $A$ to $B$."
  • Let $G$ have graph $\{(1,2), (2,4), (3,2), (5,8)\}$ (same as above), and define $A = \{1, 2, 3, 5\}$ and $C = \{2, 4, 8, 9, 11, \pi, 3/2\}$. Then $G:A\to C$ is a (admittedly ridiculous) function. Note that all the second coordinates of the graph are in $C$, along with a bunch of miscellaneous suspicious characters that are not second coordinates of pairs in the graph.
  • Let $H$ have graph $\{(1,2), (2,4), (3,2), (5,8)\}$. Then $H:A\to \mathbb{R}$ is a function, since $2$, $4$ and $8$ are all real numbers.
  • Let $D = \{1, 2, 5\}$ and $E = \{1, 2, 3, 4, 5\}$. Then there is no function $D\to A$ and no function $E\to A$ with graph $\{(1,2), (2,4), (3,2), (5,8)\}$. Neither $D$ nor $E$ has exactly the same elements as the first coordinates of the graph.

Identity and inclusion

Suppose we have two sets  A and  B with $A\subseteq B$.

  • The identity function on A is the function ${{\operatorname{id}}_{A}}:A\to A$ defined by ${{\operatorname{id}}_{A}}(x)=x$ for all $x\in A$. (Many authors call it ${{1}_{A}}$).
  • When $A\subseteq B$, the inclusion function from $A$ to $B$ is the function $i:A\to B$ defined by $i(x)=x$ for all $x\in A$. Note that there is a different function for each pair of sets $A$ and $B$ for which $A\subseteq B$. Some authors call it ${{i}_{A,\,B}}$ or $\text{in}{{\text{c}}_{A,\,B}}$.

The identity function and an inclusion function for the same set $A$ have exactly the same graph, namely $\left\{ (a,a)|a\in A \right\}$. More about this below.

Other definitions of function

Original abstract definition of function

Definition

Remarks

Possible confusion

Some confusion can result because of the presence of these two different definitions.

Multivalued function

Some older mathematical papers in com­plex func­tion theory do not tell you that their functions are multi­valued. There was a time when com­plex func­tion theory was such a Big Deal in research mathe­matics that the phrase "func­tion theory" meant complex func­tion theory and all the cogno­scenti knew that their functions were multi­valued.

The phrase multivalued function refers to an object that is like a function $f:S\to T$ except that for $s\in S$, $f(s)$ may denote more than one value.

Examples

  • Multivalued functions arose in considering complex functions. In common practice, the symbol $\sqrt{4}$ denoted $2$, although $-2$ is also a square root of $4$. But in complex function theory, the square root function takes on both the values $2$ and $-2$. This is discussed in detail in Wikipedia.
  • The antiderivative is an example of a multivalued operator. For any constant $C$, $\frac{x^3}{3}+C$ is an antiderivative of $x^2$.

A multivalued function $f:S\to T$ can be modeled as a function with domain $S$ and codomain the set of all subsets of $T$. The two meanings are equivalent in a strong sense (naturally equivalent}). Even so, it seems to me that they represent two differ­ent ways of thinking about multivalued functions. ("The value may be any of these things…" as opposed to "The value is this whole set of things.")

The phrases "multivalued function" and "partial function" upset some picky types who say things like, "But a multi­valued func­tion is not a func­tion!". A step­mother is not a mother, either. See the Hand­book article on radial category.

Partial function

A partial function $f:S\to T$ is just like a function except that its input may be defined on only a subset of $S$. For example, the function $f(x)=\frac{1}{x}$ is a partial function from the real numbers to the real numbers.

This models the behavior of computer programs (algorithms): if you consider a program with one input and one output as a function, it may not be defined on some inputs because for them it runs forever (or gives an error message).

In some texts in computing science and mathematical logic, a function is by convention a partial function, and this fact may not be mentioned explicitly, especially in research papers.

New approaches to functions

All the definitions of function given here produce mathematical structures, using the traditional way to define mathematical objects in terms of sets. Such definitions have disadvantages.

Mathematicians have many ways to think about functions. That a function is a set of ordered pairs with a certain property (functional) and possibly some ancillary ideas (domain, codomain, and others) is not the way we usually think about them$\ldots$Except when we need to reduce the thing we are studying to its absolutely most abstract form to make sure our proofs are correct. That most abstract form is what I have called the rigorous view or the dry bones and it is when that reasoning is needed that the sets-with-structure approach has succeeded.

Our practice of abstraction has led us to new approaches to talking about functions. The most important one currently is category theory. Roughly, a category is a bunch of objects together with some arrows going between them that can be composed head to tail. Functions between sets are examples of this: the sets are the objects and the functions the arrows.

This abstracts the idea of function in a way that brings out common ideas in various branches of math. Research papers in many branches of mathematics now routinely use the language of category theory. Categories now appear in some undergraduate math courses, meaning that Someone needs to write a chapter on category theory for abstractmath.org.

Besides category theory, computing scientists have come up with other abstract ways of dealing with functions, for example type theory. It has not come as far along as category theory, but has shown recent signs of major progress.

Both category theory and type theory define math objects in terms of their effect on and relationship with other math objects. This makes it possible to do abstract math entirely without using sets-with-structure as a means of defining concepts.

 

Function and codomain

I recently posted the following information in the talk page of the Wikipedia article on functions, where they were arguing about whether "function" means a set of ordered pairs with the functional property or a structure with a domain $D$, a codomain $C$, and a graph $G$ which is a subset of $D\times C$ with the functional property.

I collected data from some math books published since 2000 that contain a definition of function; they are listed below.  In this list, "typed" means  function was defined as going from a set A to a set B, A was called the domain, and B was not given a name. If "typed" is followed by a word (codomain, range or target) that was the name given the codomain. One book defined a function essentially as a partial function. Some that did not name the codomain defined "range" in the sense of image. Some of them emphasized that the range/image need not be the same as the codomain.

As far as I know, none of these books said that if two functions had the same domain and the same graph but different codomains they had to be different functions.  But I didn't read any of them extensively. 

My impression is that modern mathematical writing at least at college level does distinguish the domain, codomain, and image/range of a function, not always providing a word to refer to the codomain.

If the page number as a question mark after it that means I got the biblio data for the book from Amazon and the page number from Google books, which doesn't give the edition number, so it might be different.

I did not look for books by logicians or computing scientists.  My experience is that logicians tend to use partial functions and modern computing scientists generally require the codomain to be specified.

Opinion:  If you don't distinguish functions as different if they have different codomains, you lose some basic intuition (a function is a map) and you mess up common terminology.  For example the only function from {1} to {1} is the identity function, and is surjective.  The function from {1} to the set of real numbers (which is a point on the real line) is not the identity function and is not surjective.

THE LIST

Mathematics for Secondary School Teachers
 By Elizabeth G. Bremigan, Ralph J. Bremigan, John D. Lorch, MAA 2011
p. 6 (typed)

Oxford Concise Dictionary of Mathematics, ed. Christopher Clapham and James Nicholson,  Oxford University Press, 4th ed., 2009.
p. 184, (typed, codomain)

Math and Math-in-school: Changes in the Treatment of the Function Concept in …
 By Kyle M. Cochran, Proquest, 2011
p74  (partial function)

 Discrete Mathematics: An Introduction to Mathematical Reasoning
 By Susanna S. Epp, 4th edition, Cengage Learning, 2010 
 p. 294? (typed, co-domain)

 Teaching Mathematics in Grades 6 – 12: Developing Research-Based …
 By Randall E. Groth, SAGE, 2011
 p236 (typed, codomain)

Essentials of Mathematics, by Margie Hale, MAA, 2003.
p. 38 (typed, target).

Elements of Advanced Mathematics
 By Steven G. Krantz, 3rd ed., Chapman and Hall, 2012
p79? (typed, range)

Bridge to Abstract Mathematics
 By Ralph W. Oberste-Vorth, Aristides Mouzakitis, Bonita A. Lawrence, MAA 2012
 p76 (typed, codomain)

The Road to Reality by Roger Penrose, Knopf, 2005.
p. 104 (typed, target)

Precalculus: Mathematics for Calculus
 By James Stewart, Lothar Redlin, Saleem Watson, Cengage, 2011
p. 143.  (typed)

The Mathematics that Every Secondary School Math Teacher Needs to Know
 By Alan Sultan, Alice F. Artzt , Routledge, 2010.
 p.400 (typed)
 
 

Mathematical definitions — second attempt

I have posted a revised version of the abstractmath.org article on mathematical definitions
here. I previously attempted to post it on this blog but the addition of the necessary css file had various side effects elsewhere in my Word Press setup, so I deleted the post.

The power of being naive

To manipulate the demos in this post, you must have Wolfram CDF Player installed on your computer. It is available free from the Wolfram website. The code for the demos is in the Mathematica notebook MM Def Deriv.nb. See How to manipulate the diagrams for more information on what you can do with them.

Learning about the derivative as a concept

The derivative $f'(x)$ of $f(x)$ is the function whose value at $a$ is the slope of the line tangent to the graph $y=f(x)$ at the point $(a,f(a))$.

To gain understanding of the concept of derivative the student need to see and play with the pictures that illustrate the definition. This can be done in stages:

  • Give an intuitive, pictorial explanation of the tangent line.
  • Show in pictures what the slope of a line is.
  • Show in pictures how you can approximate the tangent line with secant lines.

Of course, many teachers and textbooks do this. I propose that:

The student will benefit in the long run by spending a whole class session on the intuitive ideas I just described and doing a set homework based only on intuition. Then you can start doing the algebraic stuff.

This post provides some ideas about manipulable diagrams that students can play with to gain intuition about derivatives. Others are possible. There are many on the Mathematica Demonstrations website. There are others written in Java and other languages, but I don't know of a site that tries to collect them in one place.

My claim that the student will benefit in the long run is not something I can verify, since I no longer teach.

Present the tangent line conceptually

The tangent line to a curve

  • is a straight line that touches the curve at a point on the curve,
  • and it goes in the same direction that the curve is going, like the red line in the picture below. (See How to manipulate the diagrams.)

 

My recommendation is that you let the students bring up some of the fine points.

  • The graph of $y=x^3-x$ has places where the tangent line cuts the curve at another point without being parallel to the curve there. Move the slider to find these places.
  • The graph of $y=\cos(\pi x)$ has places where the same line is tangent at more than one point on the curve. (This may requre stepping the slider using the incrementers.)
  • Instigate a conversation about the tangent line to a given straight line.
  • My post Tangents has other demos intended to bother the students.
  • Show the unit circle with some tangent lines and make them stare at it until they notice something peculiar.
  • "This graph shows the tangent line but how do you calculate it?" You can point out that if you draw the curve carefully and then slide a ruler around it so that it is tangent at the point you are interested in, then you can draw the tangent carefully and measure the rise and run with the ruler. This is a perfectly legitimate way to estimate the value of the slope there.

Slope of the tangent line conceptually

This diagram shows the slope of the tangent line as height over width.

  • Slide the $x$ slider back and forth. The width does not change. The height is measured from the tangent line to the corner, so the height does change; in particular, it changes sign appropriately.
  • This shows that the standard formula for the derivative of the curve gives the same value as the calculated slope of the tangent. (If you are careful you can find a place where the last decimal places differ.) You may want to omit the "derivative value" info line, but most students in college calculus already know how to calculate the formulas for the derivative of a polynomial– or you can just tell them what it is in this case and promise to show how to calculate the formula later.
  • Changing the width while leaving $x$ fixed does not change the slope of the tangent line (up to roundoff error).
  • In fact I could add another parameter that allows you to calculate height over width at other places on the tangent line. But that is probably excessive. (You could do that in a separate demo that shows that basic property that the slope of a straight line does not change depending on where you measure it — that is what a curve being a straight line means.)
  • This graph provides a way to estimate the slope, but does not suggest a way to come up with a formula for the slope, in other words, a formula for the derivative.

Conceptual calculation of the slope

This diagram shows how to calculate the value of the slope at a point using secant lines to approximate the tangent line. If you have a formula for the function, you can calculate the limit of the slope of the secant line and get a formula for the derivative.

 

  • The function $f(x)=x^3-x$.
  • The secant points are $(x-h,f(x-h))$ and $(x+h, f(x+h))$. $h$ is called "width" in the diagram.
  • Moving $x$ with the slider shows how the tangent line and secant line have similar slopes.
  • Moving the width to the left, to $0$ (almost), makes the secant line coincide with the tangent line. So intuitively the limit of the slope of the secant line is the slope of the tangent line.
  • The distance between the secant points is the Euclidean distance. (It may be that including this information does not help, so maybe it should be left out.)
  • The slope of the secant line is $\frac{f(x+h)-f(x-h)}{(x+h)-(x-h)}$ when $h\neq0$. This simplifies to $3x^2+h^2-1$, so the limit when $h\to0$ is $3x^2-1$, which is therefore a formula for the derivative function.

 

Testing intuitive concepts

Most of the work students do when studying derivatives is to solve some word problems (rate of change, maximization) in which the student is expected to come up with an appropriate function $f(x)$ and then know or find out the formula for $f'(x)$ in the process of solving the problem. In other words there is a heavy emphasis on computation and much less on concept.

The student in the past has had to do very few homework problems that test for understanding the concept. Lately some texts do have problems that test the concept, for example:

This is the graph of a function and its derivative. Which one is the function and which is its derivative?

Concept Prob

Note that the problem does not give you the formula for the function, nor does it have to.

Many variations are possible, all involving calculating parameters directly from the graph:

  • "These are the first and second derivatives of a function. Where (within the bounds of the graph) is the function concave up?"
  • "These are the first and second derivatives of a function. Where (within the bounds of the graph) are its maxima and minima?"
  • "This straight line is the derivative of a function. Show that the function is a quadratic function and measure the slope of the line in order to estimate some of the coefficients of the quadratic."

 

How to manipulate the diagrams

 

  • You can move the sliders back and forth to to move to different points on the curve.
  • In the first diagram, you can click on one of the four buttons to see how it works for various curves.
  • The arrow at the upper right makes it run automatically in a not very useful sort of way.
  • The little plus sign below the arrow opens up some other controls and a box showing the value of $a$, including step by step operation (plus and minus signs).
  • If you are using Mathematica, you can enter values into the box, but if you are using CDF Player, you can only manipulate the number using the slider or the plus and minus incrementers.

 

Monads for High School III: Algebras

This is a continuation of Monads for high school I and Monads for High School II: Lists. This post covers the concept of algebras for the monad for lists.

To manipulate the demos in this post, you must have Wolfram CDF Player installed on your computer. It is available free from the Wolfram website. The code for the demos is in the Mathematica notebook MonadAlg.nb

Lists

$\textrm{Lists}(S)$ is the set of all lists of finite length whose entries are elements of $S$.

  • $\boxed{2\; 2\; 4}$ is the way I denote the list of length $3$ whose first and second entries are each $2$ and whose third entry is $4$.
  • A list with only one entry, such as $\boxed{2}$, is called a singleton list.
  • The empty list $\boxed{\phantom{2}}$ has no entries.
  • $\textrm{Lists}^*(S)$ is the set of all nonempty lists of finite length whose entries are elements of $S$.
  • $\textrm{Lists}(\textrm{Lists}(S))$ is the list whose entries are lists with entries from $S$.
  • For example, $\boxed{\boxed{5\; 7}\; \boxed{2\; 12\; 7}}$ and $\boxed{\boxed{5\; 7\; 2\; 12\; 7}}$ are both entries in $\textrm{Lists}^*(\textrm{Lists}^*(\mathbb{Z}))$. The second one is a singleton list!
  • $\boxed{\boxed{\phantom{3}}\; \boxed{2}}
    $ and $\boxed{\boxed{\phantom{3}}}$ are entries in $\textrm{Lists}^*(\textrm{Lists}(\mathbb{Z}))$.
  • The empty list $\boxed{\phantom{2}}$ is an entry in $\textrm{Lists}(\mathbb{Z})$, in $\textrm{Lists}(\textrm{Lists}^*(\mathbb{Z}))$ and in $\textrm{Lists}(\textrm{Lists}(\mathbb{Z}))$. If you have stared at this for more than ten minutes, do something else and come back to it later.

The star notation is used widely in math and computing science to imply that you are including everything except some insignificant shrimp of a thing such as the empty list, the empty set, or $0$. For example, $\mathbb{R}^*$ denotes the set of all nonzero real numbers.

More details about lists are in Monads for High School II: Lists.

Join

The function join (or concatenation) takes two lists and creates a third list. For example, if you join $\boxed{5\; 7}$ to $\boxed{2\; 12\; 7 }$ in that order you get $\boxed{5\; 7\; 2\; 12\; 7}$.

  • I will use this notation: join$\boxed{\boxed{5\; 7}\; \boxed{2\; 12\; 7}}=\boxed{5\; 7\; 2\; 12\; 7}$.
  • This notation means that I am regarding join as a function that takes a two-element list in $\textrm{Lists}(\textrm{Lists}(S))$ to an element of $\textrm{Lists}(S)$.
  • join removes one level of lists
  • join is not commutative: join$\boxed{\boxed{2\; 12\; 7}\; \boxed{5\; 7}}=\boxed{2\; 12\; 7\; 5\; 7}$
  • Join is associative, and as for any associative binary operation, join is defined on any finite list of lists of elements of $S$. So for example, join$\boxed{\boxed{5\; 7}\; \boxed{2\; 12\; 7}\; \boxed{1}}=\boxed{5\; 7\; 2\; 12\; 7\; 1}$.
  • For any single list $\boxed{a\; b\; c}$, join$\boxed{\boxed{a\; b\; c}}=\boxed{a\; b\; c}$. This is required to make the theory work. It is called the oneidentity property.
  • If the empty list $\boxed{\phantom{2}}$ occurs in a list of lists, it disappears when join is applied: join $\boxed{\boxed{2\; 3}\; \boxed{\phantom{2}}\; \boxed{4\; 5\; 6}}=\boxed{2\; 3\; 4\; 5\; 6}$.

More details about join in Monads for High School II: Lists.

The main monad diagram

When you have a list of lists of lists, join can be applied in two different ways, "inside" and "outside" as illustrated in the diagram below. It gives you several different inputs to try out as a way to understand what is happening.

This is the special case of the main diagram for all monads as it applies to the List monad.

As you can see, after doing either of "inside" and "outside", if you then apply join, you get the same list. That list is simply the list of entries in the beginning list (and the two intermediate ones) in the same order, disregarding groupings.

From what I have just written, you must depend on your pattern recognition abilities to learn what inside and outside mean. But both can also be described in words.

  • The lists outlined in black are lists of elements of $\mathbb{Z}$. In other words, they are elements of $\textrm{Lists}(\mathbb{Z})$.
  • The lists outlined in blue are lists of elements of $\textrm{Lists}(\mathbb{Z})$. In other words, they are list of lists of elements of $\mathbb{Z}$. Those are the kinds of things you can apply join to.
  • The leftmost list in the diagram, outlined in green, is a list in $\textrm{Lists}(\textrm{Lists}(\mathbb{Z}))$. This means you can apply join in two different ways:
  • Each list boxed in blue is a list of lists of integers (two of the are singletons!) so you can apply join to each of them. This is joining inside first.
  • You can apply join directly to the leftmost list, which is a list of lists (of lists, but forget that for the moment), so you can apply join to the blue lists. This is join outside first.

To understand this diagram, staring at the diagram (for most people) uses the visual pattern recognition part of your brain (which uses over a fifth of the energy used by your brain) to understand what inside and outside mean, and then check your understanding by reading the verbal description. Starting by reading the verbal description first does not work as well for most people.

The unit monad diagram

There is a second unitary diagram for all monads:

The two right hand entries are always the same. Again, I am asking you to use your pattern recognition abilities to learn what singleton list and singleton each mean.

The main and unit monad diagrams will be used as axioms to give the general definition of monad. To give those axioms, we also need the concepts of functor and natural transformation, which I will define later after I have finished the monad algebra diagrams for Lists and several other examples.

Algebras for the List monad

If you have any associative binary operation on a set $S$, its definition can be extended to any nonempty list of elements (see Monads for High School I.)

Plus and Times are like that:

  • $(3+2)+4$ and $3+(2+4)$ have the same value $9$, so you can write $3+2+4$ and it means $9$ no matter how you calculate it.
  • I will be using the notation Plus$\boxed{3\; 2\; 4}$ instead of $3+2+4$.
  • Times is also associative, so for example we can write Times$\boxed{3\; 2\; 4}=24$.
  • Like join, we require that these operations satisfy oneidentity, so we know Plus$\boxed{3}=3$ and Times$\boxed{3}=3$.
  • When the associative binary operation has an identity element, you can also define its value on the empty list as the identity element: Plus$\boxed{\phantom{3}}=0$ and Times$\boxed{\phantom{3}}=1$. I recommend that you experiment with examples to see why it works.

An algebra for the List monad is a function algop:$\textrm{Lists}(S)\to S$ with certain properties: It must satisfy the Main Monad Algebra Diagram and the Unit Monad Algebra Diagram, discussed below.

The main monad algebra diagram

Example using Plus and Times

The following interactive diagram allows you to see what happens with Plus and Times. Afterwards, I will give the general definition.

Plus insides replaces each inside list with the result of applying Plus to it, and the other operation Join is the same operation I have used before.

Another example

The main monad algebra diagram requires that if you have a list of lists of numbers such as the one below, you can add up each list (Plus insides) and then add up the list of totals (top list in diagram), you must get the same answer that you get when you join all the lists of numbers together into one list (bottom list in the diagram) and then add up that list.

This is illustrated by this special case of the main monad algebra diagram for Plus:

General statement of the main monad algebra diagram

Suppose we have any function $\blacksquare$ $:\textrm{Lists}(S)\to S$ for any set $S$.
If we want to give the main monad algebra diagram for $\blacksquare$ we have a problem. We know for example that Plus$\boxed{1\; 2}=3$. But for some elements $a $ and $b$ of $S$, we don’t know what $\blacksquare\boxed{a\; b}$ is. One way to write it is simply to write $\blacksquare\boxed{a\; b}$ (the usual way we write a function). Or we could use tree notation and write

newalopdouble.

I will use tree notation mostly, but it is a good exercise to redraw the diagrams with functional notation.

Main monad diagram in prose

Below is a presentation of the general main monad algebra diagram using (gasp!) English phrases to describe the nodes.

genalgdiag

The unit monad algebra diagram

Suppose $\blacksquare$ is any function from $\textrm{Lists}(S)$ to $S$ for any set $S$. Then the diagram is

UnitMAdiag

This says that if you apply $\blacksquare$ to a singleton you get the unique entry of the singleton. This is not surprising: I defined above what it means when you apply an operation to a singleton just so this would happen!

A particular example

These are specific examples of the general main monad algebra diagram for an arbitrary operation $\blacksquare$:

stalgdiagleft

staldiagright

These examples show that if $\blacksquare$ is any function from $\textrm{Lists}(S)$ to $S$ for any set $S$, then

newalopleft

equals

newaloptriple

and

newalopright

equals

newaloptriple

Well, according to some ancient Greek guy, that means

newalopleft

equals

newalopright

which says that
newalopdouble
is an associative binary operation!

The mother of all associative operations

We also know that any associative binary $\blacksquare$ on any set $S$ can be extended to a function on all finite nonempty lists of elements of $S$. This is the general associative law and was discussed (without using that name) in Monads fo High School I.

Let’s put what we’ve done together into one statement:

Every associative binary operation $\blacksquare$ on a set $S$ can be extended uniquely to a function $\blacksquare:\textrm{Lists}^*(S)\to S$ that satisfies both the main monad algebra diagram and the unit monad algebra diagram. Furthermore, any function $\blacksquare:\textrm{Lists}^*(S)\to S$ that satisfies both the main monad algebra diagram and the unit monad algebra diagram is an asssociative binary operation when applied to lists of length $2$ of elements of $S$.

That is why I claim that the NonemptyList monad is the mother of all associative binary operations.

I have not proved this, but the work in this and preceding posts provide (I think) a good intuitive understanding of this fundamental relationship between lists and associative binary operations.

Things to do in upcoming posts

  • I have to give a proper definition of monads using the concepts of functor and natural transformation. I expect to do this just for set functors, not mentioning categories.
  • Every type of binary operation that is defined by equations corresponds to a monad which is the mother of all binary operations of that type. I will give examples, but not prove the general case.

Other examples of monads

  • Associative binary operations on $S$ with identity element (monoids) corresponds to all lists, including the empty list, with entries from $S$.
  • Commutative, associative and idempotent binary operations, like and and or in Boolean algebra, correspond to the set monad: $\text{Sets}(S)$ is the set of all finite and countably infinite sets of elements of $S$. (You can change the cardinality restrictions, but you have to have some cardinality restrictions.) Join is simply union.
  • Commutative and associative binary operations corresponds to the multiset monad (with a proper definition of join) and appropriate cardinality restrictions. You have to fuss about identity elements here, too.
  • Various kinds of nonassociative operations get much more complicated, involving tree structures with equivalence relations on them. I expect to work out a few of them.
  • There are lots of monads in computing science that you never heard of (unless you are a computing scientist). I will mention a few of them.

  • Every type of binary operation defined by equations corresponds to a monad. But some of them are unsolvable, meaning you cannot describe the monad precisely.

There will probably be long delay before I get back to this project. There are too many other things I want to do!