EVIL terminology

EVIL terminology

[Juergen Koslowski]

Dear all,

In photo circles EVIL is currently used as an acronym for "Electronic
Viewfinder Interchangable Lens" cameras, a new form of cameras that do
away with the optical viefinder.  Probably this is as firmly
toungue-in-cheek as the use of the term "evil" in certain areas of
category theory.

And while I don't know if many photographers are offended by this term,
this obviously is the case with some of our colleagues in category
theory.  While I'm sure that everybody agrees that the concept as such
is important and worth studying, the ongoing dispute over the
terminology seems to be rather counterproductive.

Of course, one problem with mathematidcal terminology is that so many
terms are overloaded by now (natural, continuous...)  Any term borrowed
from the realm of everyday life will carry certain overtones (and will
irritate non-mathematicians).  On the other hand, acronyms can come
across a clumsy and an impediment to speach (ufl-functors).

Maybe it is not too late to find a clever new name for the concept that
does convey the negative connotations intended by the original adopters
of "evil" without being so offensive to others (we don't want to wait
30 or more years before changing terminology, cf. the 2005/06 discussion
concerning "cartesian" and "cocartesian").  Peter Freyd was always very
inventive coining new terminology, even though not much of it caught on.

Let me start by proposing "equi-unstable" and "precarious" as possible
replacements for "evil".

-- Juergen

-- 
Juergen Koslowski               If I don't see you no more on this world
ITI, TU Braunschweig               I'll meet you on the next one
koslowj@iti.cs.tu-bs.de               and don't be late!
http://www.iti.cs.tu-bs.de/~koslowj      Jimi Hendrix (Voodoo Child, SR)


[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

Re: potential names

[Martin Escardo]

What about "dangerous" rather than "evil"? (I also dislike religious and 
moral terminology for mathematics. I specially dislike the terminology 
"morally true", which I often hear, although I like the concept.)

We frequently encounter dangerous situations in our mathematical paths, 
no matter how hard we try to avoid them, but if we know what we are 
doing it is ok.

In this discussion about "evilness", what is important is to know that 
lack of invariance under isomorphisms is dangerous (and why this is so), 
and that you must consciously know how to properly proceed if your 
definition happens to be not invariant, be it for necessity, lack of a 
better idea, or mere convenience.

MHE

Joyal wrote:
> Dear Juergen,
> 
> You wrote:
> 
>> Let me start by proposing "equi-unstable" and "precarious" as possible
>> replacements for "evil".
> 
> I agree that we should explore various options.
> Strickly speaking, there is no need for a new terminology
> since "non-invariant" has been used in the past
> and it is working pretty well.
> 
> Let me draw a list of potential names.
> I have added a few:
> 
> non-invariant
> equi-unstable
> precarious
> unstable
> fragile
> private
> 
> I invite everyone to contribute to the list.
> 
> Best,
> AJ
> 

[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

Re: potential names

[Todd Trimble]

Of the choices offered here, I like "precarious", "unstable", or
"fragile".  I just thought of "risky" myself.  Good, experienced
mathematicians will know when it's okay to take "risks" (and
will be aware of what the risks are).

"Unstable" seems like a very sober choice, not too likely to
ruffle feathers.

I encourage anyone who cares about this issue to take a look
at the current discussion at the nForum:

http://www.math.ntnu.no/~stacey/Mathforge/nForum/comments.php?DiscussionID=1886

You will see that opinions are divided even among workers in
the nLab.

Todd

----- Original Message ----- 
From: "Joyal, André" <joyal.andre@uqam.ca>
To: "Juergen Koslowski" <koslowj@iti.cs.tu-bs.de>
Sent: Tuesday, September 28, 2010 9:57 AM
Subject: categories: potential names


Dear Juergen,

You wrote:

>Let me start by proposing "equi-unstable" and "precarious" as possible
>replacements for "evil".

I agree that we should explore various options.
Strickly speaking, there is no need for a new terminology
since "non-invariant" has been used in the past
and it is working pretty well.

Let me draw a list of potential names.
I have added a few:

non-invariant
equi-unstable
precarious
unstable
fragile
private

I invite everyone to contribute to the list.

Best,
AJ


[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

potential names

[Paul Taylor]

Todd Trimble wrote:
> Of the choices offered here, I like "precarious", "unstable", or
> "fragile".  I just thought of "risky" myself.  Good, experienced
> mathematicians will know when it's okay to take "risks" (and
> will be aware of what the risks are).
>
> "Unstable" seems like a very sober choice, not too likely to
> ruffle feathers.

Mathematical terminology should not employ words that are merely
value judgements, without relevant content.  It doesn't make any
difference whether they are offensive or inoffensive value-judgement
words.  This is not how we should choose scientific terminology.

We are already cursed with vast over-use of the words "regular"
and "normal" in mathematics.   Roughly translated, these mean
"the objects that I want to study" - other people may have very
good reasons for studying other kinds of objects.

("Stable" and "sober" already have several meanings.)

There is a problem here in that there is nothing in the education
of a pure mathematician that teaches how to make a professional
judgement.   I never thought I would find myself defending
software engineering (the religion whose creed it is that programs
are better if their authors wear suits, draw diagrams and attend
committee meetings) but when computer science students are subjected
to this at least they learn that, whatever they do, they are making
professional judgements.

Since pure mathematicians do nothing similar in their training
they are easily mis-led by the use of terminology that is based
on value judgements.   They just think that they are taking
dictation from God.

Even if there is a very strong argument in favour of a particular
value judgement (as there may well be in the case under discussion)
we should still not use words that have no other content, simply
because we will want to make OTHER value judgements in future.

The English language reportedly now contains over a million words.
Can you really not find anything in this vast thesaurus (=treasury)
that describes the situation more appropriately and precisely?
There is less, not more, of an excuse if you speak French, Spanish
or another language: English allows almost completely free
immigration of words.

Let's have a bit of imagination with language, please.

Despite the abuse that I received for it here, I am rather pleased
with my introduction of the words "prone" and "supine" for the two
different orthogonal notions to "vertical" in a fibration.

A word was needed to replace "open" for an object whose terminal
projection is an open map, since subobjects with this property
have a habit of being closed subspaces.   The rich English
vocabulary offered "overt", which means "explicit".   Since I first
used this word, it has emerged that this idea is very closely related
to recursive enumerability, ie to having an explicit presentation,
so this has turned out to be a very good choice of word.

On the other hand, I regret introducing "bilimit" and "bifinite"
in domain theory.

In the case under discussion we need to distinguish between equal
and isomorphic objects.   In existing terminology, a category in
which isomorphic objects are equal is called "skeletal", although
I doubt whether this word ever gets another outing after the
definition of a category and basic concepts therein has been
given for the first time to students.

So my suggestion is that you play around with skeletons, bones
or even the Grim Reaper for something more suitable.

Paul Taylor



[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

Re: potential names

[Vaughan Pratt]

On 10/2/2010 10:43 AM, Paul Taylor wrote:
> So my suggestion is that you play around with skeletons, bones
> or even the Grim Reaper for something more suitable.

Along those lines, how about "scythe" for the functor [-,1+1]: C^op -->
C?  In a Boolean topos this is just the (internal) contravariant power
object functor, but more generally it tends to cleave objects into their
connected components in a suitable sense.

Vaughan


[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

Re: potential names

[Todd Trimble]

Paul Taylor wrote:

> Mathematical terminology should not employ words that are merely
> value judgements, without relevant content.

Does "equi-unstable" have enough relevant content for you?
Or do any of the proposals so far have enough relevant content
for you?  (I mean the pithy proposals contained in Joyal's list;
there is always the longer "non-invariant under equivalence".)

The quotation above seems to effectively summarize what the
rant is all about; I think I can do without the judgment or
projection that I or anyone else is taking "dictation from God".

> Let's have a bit of imagination with language, please.

Yes, let's. I look forward to other suggestions as well.

> So my suggestion is that you play around with skeletons, bones
> or even the Grim Reaper for something more suitable.

Thanks for the suggestion, but I don't much care for it
if it merely evokes skeletal subcategories -- that is only
one application of the concept we're discussing.

Do you have a positive contribution you'd like to make,
Paul?

Todd

----- Original Message -----
From: "Paul Taylor" <pt10@PaulTaylor.EU>
To: "Categories list" <categories@mta.ca>
Sent: Saturday, October 02, 2010 1:43 PM
Subject: categories: potential names


> Todd Trimble wrote:
>> Of the choices offered here, I like "precarious", "unstable", or
>> "fragile".  I just thought of "risky" myself.  Good, experienced
>> mathematicians will know when it's okay to take "risks" (and
>> will be aware of what the risks are).
>>
>> "Unstable" seems like a very sober choice, not too likely to
>> ruffle feathers.
>
> Mathematical terminology should not employ words that are merely
> value judgements, without relevant content.  It doesn't make any
> difference whether they are offensive or inoffensive value-judgement
> words.  This is not how we should choose scientific terminology.
>
> We are already cursed with vast over-use of the words "regular"
> and "normal" in mathematics.   Roughly translated, these mean
> "the objects that I want to study" - other people may have very
> good reasons for studying other kinds of objects.
>
> ("Stable" and "sober" already have several meanings.)
>
> There is a problem here in that there is nothing in the education
> of a pure mathematician that teaches how to make a professional
> judgement.   I never thought I would find myself defending
> software engineering (the religion whose creed it is that programs
> are better if their authors wear suits, draw diagrams and attend
> committee meetings) but when computer science students are subjected
> to this at least they learn that, whatever they do, they are making
> professional judgements.
>
> Since pure mathematicians do nothing similar in their training
> they are easily mis-led by the use of terminology that is based
> on value judgements.   They just think that they are taking
> dictation from God.
>
> Even if there is a very strong argument in favour of a particular
> value judgement (as there may well be in the case under discussion)
> we should still not use words that have no other content, simply
> because we will want to make OTHER value judgements in future.
>
> The English language reportedly now contains over a million words.
> Can you really not find anything in this vast thesaurus (=treasury)
> that describes the situation more appropriately and precisely?
> There is less, not more, of an excuse if you speak French, Spanish
> or another language: English allows almost completely free
> immigration of words.
>
> Let's have a bit of imagination with language, please.
>
> Despite the abuse that I received for it here, I am rather pleased
> with my introduction of the words "prone" and "supine" for the two
> different orthogonal notions to "vertical" in a fibration.
>
> A word was needed to replace "open" for an object whose terminal
> projection is an open map, since subobjects with this property
> have a habit of being closed subspaces.   The rich English
> vocabulary offered "overt", which means "explicit".   Since I first
> used this word, it has emerged that this idea is very closely related
> to recursive enumerability, ie to having an explicit presentation,
> so this has turned out to be a very good choice of word.
>
> On the other hand, I regret introducing "bilimit" and "bifinite"
> in domain theory.
>
> In the case under discussion we need to distinguish between equal
> and isomorphic objects.   In existing terminology, a category in
> which isomorphic objects are equal is called "skeletal", although
> I doubt whether this word ever gets another outing after the
> definition of a category and basic concepts therein has been
> given for the first time to students.
>
> So my suggestion is that you play around with skeletons, bones
> or even the Grim Reaper for something more suitable.
>
> Paul Taylor
>

[For admin and other information see: http://www.mta.ca/~cat-dist/ ]

Re: potential names

[Fred E.J. Linton]

Following up on Paul Taylor's suggestion to " ... play around with skeletons,

bones or even the Grim Reaper ... ", Vaughan Pratt <pratt@cs.stanford.edu> 
offered:

> ... how about "scythe" for the functor [-,1+1]: C^op -->
> C?  In a Boolean topos this is just the (internal) contravariant power
> object functor, but more generally it tends to cleave objects into their
> connected components in a suitable sense.

But "scythe"? A scythe is for cutting off stalks, wholesale.
(A sickle is for cutting them off ... umm ... retail, perhaps :-) .)

To "cleave" an object in two one might utilize a *cleaver*, or, more 
bluntly, a *wedge* -- but the resulting two parts are probably *not*
in most instances "connected" at all, let alone "connected components":
rather, they merely provide a binary coproduct decomposition
of the given object.

Cheers, -- Fred



[For admin and other information see: http://www.mta.ca/~cat-dist/ ]