On 09 Jun 2007 00:22:14 -0400, Kent M Pitman <······@nhplace.com> said:
| Vassil Nikolov <···············@pobox.com> writes:
|| Is there any terminology to distinguish different kinds of ATOMs in
|| lisp, e.g. characters, numbers and symbols, etc. vs. arrays and
|| structures, etc.? Or a term that would group arrays, conses,
|| structures, etc. into one category?
|| ...
| On an informal basis, I sometimes use the term "container" to talk
| about the union of the types cons, vector, sequence, hash table,
| etc. (i.e., the types that have subparts).
Thanks, "container" answers one of my questions.
(I have also been thinking of using "molecule" for this same
purpose, but that may be too pretentious.)
| I don't think I know a term that partitions only the atoms in this
| way. That's kind of a strange set, in part because the line between
| atoms and non-atoms makes very sense as meaning anything other than
^^^^^^^^^^ I read this as "very little sense"
| non-cons. So if liked "container" as the name for the larger set,
^^^^^^^^ "if you liked"?
| your question (literally taken, at least) asks for a name for the
| non-cons containers.
Or, better yet, for the non-container atoms. One possibility is the
term Barry Margolin used, "singleton". (Again along the lines of
the analogy to sciences, I have been thinking of "particle"
("elementary particle" being too long), which again may be too
pretentious, and "quark" is certainly too pretentious.)
I'm sure that we ought to note the fact that there is no popular
terminology as I am looking for here, but my thoughts on the moral
of this observation are too raw to post.
| Note that NIL is problematic in any case, being under some views an
| empty atomic container... although if you use the "no subparts"
| definition, at least NIL has a uniform interpretation whether you view
| it as a list or a symbol or a boolean.
Right.
| If the atom division is something that bugs you a lot, one way to
| address it is never to use atom and always to use the type CONS or the
| function CONSP (negated, of course) in situations where you want this
| distinction.
I have gotten used to this idiosyncratic meaning of ATOM, and then I
have no problem using the existing set of type specifiers or
predicates in programs; what I am looking for is terminology to use
in (English) descriptions of algorithms and programs.
---Vassil.
--
The truly good code is the obviously correct code.
From: Kent M Pitman
Subject: Re: terminology for subcategories of ATOM?
Date:
Message-ID: <u8xasy4sq.fsf@nhplace.com>
Vassil Nikolov <···············@pobox.com> writes:
> (I have also been thinking of using "molecule" for this same
> purpose, but that may be too pretentious.)
If it comes to going in that direction, you might also consider
"compound", since it's a simpler name with a similar connotation.
Kent M Pitman wrote:
> Vassil Nikolov <···············@pobox.com> writes:
>
>> (I have also been thinking of using "molecule" for this same
>> purpose, but that may be too pretentious.)
>
> If it comes to going in that direction, you might also consider
> "compound", since it's a simpler name with a similar connotation.
What about scalar and composite?
From: Kent M Pitman
Subject: Re: terminology for subcategories of ATOM?
Date:
Message-ID: <umyz72xes.fsf@nhplace.com>
Damien Kick <·····@earthlink.net> writes:
> Kent M Pitman wrote:
> > Vassil Nikolov <···············@pobox.com> writes:
> >
> >> (I have also been thinking of using "molecule" for this same
> >> purpose, but that may be too pretentious.)
> > If it comes to going in that direction, you might also consider
> > "compound", since it's a simpler name with a similar connotation.
>
> What about scalar and composite?
You're presumably explicitly asking for a subjective opinion, so I'll
oblige you, with the caveat that my response intends on sense of
authority, merely my own personal aesthetic judgment, as one person's
opinion.
I can see why you'd want to use these two words, but I will here tell
you why I would not, to the best I can explain such an inherently
subjective assessment:
Some people consider scalar to refer to numbers, and not necessarily
other non-subdividable things, such as characters or symbols. (The
word "atom" should have been the right term, other than due to the
historical accident of not annexing arrays, etc. as non-atoms.) I
like to avoid words that have obvious "snags" in them which are likely
to lead to serious misunderstandings, and this is strong enough that
it bugs me. I don't assert that others would necessarily agree,
though; I have no data on that.
As to composite, it's funny that as I try to work through why it feels
like not a good word, it may be that different people just hear it
differently in spite of the fact that a more reasoned look at its
etymology suggests that what I hear may not be the history of how the
word came about. (That is, it's from the root word componere, meaning
to put together or compose, yet I don't feel its modern usage is just
that, but rather that additional connotation has been added such that
I don't sense that a composite is the same kind of composed thing as a
compound is.) When I hear compound, I think of something like Lego
building blocks that has discrete components snapped together, and
that's like the thing we're talking about. When I hear composite, I
don't think of something composed of easily identifiable components
but rather two non-discrete source materials blurred together (like
the way paint mixes colors or the way eggs mix in an omelette ... or
the way atoms mix in an alloy [a statistical blur but not a new
structure], as distinguished from the way they mix in a compound [a
highly rigid pattern of dictated shape, composing a new object with
distinct identity and nature]). Now maybe that's just a personal
misunderstanding I have about the sense of that word and not something
that is really there in the language ... or maybe it's that it has
two different dialectal senses, and I have one and you have another.
A quick double-check of webbed sources on mixtures and alloys seems to
support the notion of mine that composite is more synonymous with
mixture than compound, since some sources cite synonyms among an
equivalence class like alloy, blend, composite, fusion. Others do
seem to also place a larger list of things together, to include adding
combination and compound to the list of synonyms for mixture, which I
have to say, at risk of blurring this whole combination at the
meta-level, sounds like it just mixes things up to me. But perhaps
this is because synonym lists contain too little guidance. I like it
better when a dictionary goes out of its way to give comparative
senses, and one such can be found at dictionary.com, when it refers
through to the American Heritage Dictionary of the English Language
and offers a very nice usage summary, a small portion of which I'll
quote here:
"A compound constitutes a new and independent entity: The school's
program is a compound of scholarship and athleticism.
A composite has components that may retain part of their identities:
a musical suite that is a composite of operatic themes."
-- The American Heritage(R) Dictionary of the English Language,
Fourth Edition, Copyright (C) 2006 by Houghton Mifflin Company.
Published by Houghton Mifflin Company. All rights reserved.
as viewed through http://dictionary.reference.com/browse/mixture
Kent M Pitman wrote:
> Some people consider scalar to refer to numbers, and not necessarily
> other non-subdividable things, such as characters or symbols. (The
> word "atom" should have been the right term, other than due to the
> historical accident of not annexing arrays, etc. as non-atoms.) I
> like to avoid words that have obvious "snags" in them which are likely
> to lead to serious misunderstandings, and this is strong enough that
> it bugs me. I don't assert that others would necessarily agree,
> though; I have no data on that.
Scalars are definitely numbers and only numbers, IMO. Some might even
argue that they are measurable physical quantities and therefore real
numbers (some measurable quantities can be zero or negative, assuming
some kind of relative scale). As a numerical linear algebra person, I
tend to think of scalars as objects for which the standard arithmetic
operations {+, -, *, /} make sense (assuming the usual extensions to
make "/" closed).
I personally call tensors of any dimension, lists, sets, and the like
"container types" or "aggregate types."
mfh
Kent M Pitman wrote:
> Some people consider scalar to refer to numbers, and not necessarily
> other non-subdividable things, such as characters or symbols. (The
> word "atom" should have been the right term, other than due to the
> historical accident of not annexing arrays, etc. as non-atoms.) I
> like to avoid words that have obvious "snags" in them which are likely
> to lead to serious misunderstandings, and this is strong enough that
> it bugs me. I don't assert that others would necessarily agree,
> though; I have no data on that.
Scalars are definitely numbers and only numbers, IMO. Some might even
argue that they are measurable physical quantities and therefore real
numbers (some measurable quantities can be zero or negative, assuming
some kind of relative scale). As a numerical linear algebra person, I
tend to think of scalars as objects for which the standard arithmetic
operations {+, -, *, /} make sense (assuming the usual extensions to
make "/" closed).
I personally call tensors of any dimension, lists, sets, and the like
"container types" or "aggregate types."
mfh
On Sun, 10 Jun 2007 17:38:37 -0700, Mark Hoemmen <············@gmail.com> said:
| Kent M Pitman wrote:
|| Some people consider scalar to refer to numbers, and not necessarily
|| other non-subdividable things, such as characters or symbols. (The
|| word "atom" should have been the right term, other than due to the
|| historical accident of not annexing arrays, etc. as non-atoms.) I
|| like to avoid words that have obvious "snags" in them which are likely
|| to lead to serious misunderstandings, and this is strong enough that
|| it bugs me. I don't assert that others would necessarily agree,
|| though; I have no data on that.
| Scalars are definitely numbers and only numbers, IMO. Some might even
| argue that they are measurable physical quantities and therefore real
| numbers (some measurable quantities can be zero or negative, assuming
| some kind of relative scale). As a numerical linear algebra person, I
| tend to think of scalars as objects for which the standard arithmetic
| operations {+, -, *, /} make sense (assuming the usual extensions to
| make "/" closed).
Broadly, yes, but strictly speaking, it is even more restricted than
that, as "scalar" only has a meaning in the context of linear vector
spaces, referring to the objects that, well, scale vectors by
multiplying them. Scalars are only required to form a ring (i.e.,
division is not necessarily defined), even though fields are used in
the most frequent cases, such as vector spaces over the field of the
real numbers or the field of the complex numbers. Note that the
notion of scalar is not an absolute one (for example, the elements
of R, the set of real numbers, are both scalars and vectors when R
is considered as a one-dimensional vector space over R).
In any case, the above only makes the case stronger that characters
and symbols are not "scalars" indeed (I seem to recall that the
Pascal programming language introduced such terminology; it would
have been better if it hadn't).
With regards to "atom", it's funny how history repeated itself:
atoms in the physical sense also turned out not to be indivisible
after all...
---Vassil.
--
The truly good code is the obviously correct code.
Vassil Nikolov wrote:
> On Sun, 10 Jun 2007 17:38:37 -0700, Mark Hoemmen <············@gmail.com> said:
>
> | Scalars are definitely numbers and only numbers, IMO. [...]
> | As a numerical linear algebra person, I tend to think of scalars as
> | objects for which the standard arithmetic operations {+, -, *, /} make
> | sense (assuming the usual extensions to make "/" closed).
>
> Broadly, yes, but strictly speaking, it is even more restricted than
> that, as "scalar" only has a meaning in the context of linear vector
> spaces, referring to the objects that, well, scale vectors by
> multiplying them.
Well, sure, if you ask a /mathematician/ but why would anybody want to
do that? We have already learned from a previous c.l.lisp thread that
programming has nothing to do with mathematics.
scalar. Dictionary.com. Dictionary.com Unabridged (v 1.1). Random House,
Inc. http://dictionary.reference.com/browse/scalar (accessed: June 11,
2007).
"ladderlike in arrangement or organization; graduated: a scalar
structure for promoting personnel."
Definitely not mathematics. But here is my favorite. <voice
of="Stephen Colbert">Damn you, Larry Wall!</voice>
"Any data type that stores a single value (e.g. a number or Boolean), as
opposed to an aggregate data type that has many elements. A string is
regarded as a scalar in some languages (e.g. Perl) and a vector of
characters in others (e.g. C)."
He's gone and infected Dictionary.com with his misuse! I knew I should
have never trusted anyone who would deliver a State of the Onion
address. I knew something stank about that <groan>.
On Tue, 12 Jun 2007 05:43:01 GMT, Damien Kick <·····@earthlink.net> said:
| ...
| scalar. Dictionary.com. Dictionary.com Unabridged (v 1.1). Random
| House, Inc. http://dictionary.reference.com/browse/scalar (accessed:
| June 11, 2007).
| "ladderlike in arrangement or organization; graduated: a scalar
| structure for promoting personnel."
| Definitely not mathematics.
Well, I beg to differ. Multiplication by a scalar promotes (or
demotes) a vector along a ladder collinear with that vector, so this
is not far removed from mathematics, in my opinion.
But, of course, we had not decided on a definition of mathematics
beforehand...
By the way, FWIW the American Heritage Dictionary gives a definition
completely different from the one quoted above, in part: "a quantity
... that is completely specified by its magnitude and has no
direction" (and the definition as an adjective is explicitly tagged
as a mathematical term).
| ...
| "Any data type that stores a single value (e.g. a number or Boolean),
Giving these two as examples for such a data type is somewhat of a
compromise, though (if that is the word I want), since both numbers
and Booleans can serve as scalars for a linear vector space (though
for Booleans that would be a little unusual).
| as opposed to an aggregate data type that has many elements. A string
| is regarded as a scalar in some languages (e.g. Perl) and a vector of
| characters in others (e.g. C)."
Aye, there's the rub, but in any case we are discussing lisp
terminology here, not Perl or C.
(Yes, I understand that I am following up to a facetious post.)
---Vassil.
--
The truly good code is the obviously correct code.
On 10 Jun 2007 19:38:03 -0400, Kent M Pitman <······@nhplace.com> said:
| ...
| dictionary.com, when it refers
| through to the American Heritage Dictionary of the English Language
| and offers a very nice usage summary, a small portion of which I'll
| quote here:
| "A compound constitutes a new and independent entity: The school's
| program is a compound of scholarship and athleticism.
| A composite has components that may retain part of their identities:
| a musical suite that is a composite of operatic themes."
| -- The American Heritage(R) Dictionary of the English Language,
| Fourth Edition, Copyright (C) 2006 by Houghton Mifflin Company.
| Published by Houghton Mifflin Company. All rights reserved.
| as viewed through http://dictionary.reference.com/browse/mixture
It is perhaps interesting to note that the previous edition has (in
the Synonyms section of the article for "mixture"):
A _compound_ is a combination of elements or parts that together
constitute a new and independent entity:
The word _houseboat_ is a compound. Creative genius is a
compound made up of exceptional intellect and superior
imagination.
A _composite_ usually lacks the unity of a compound since the
components may not wholly lose their identities:
The suite is a composite of themes for various parts of the
opera.
[The American Heritage Dictionary of the English Language, Third
Edition, Copyright (C) 1992 by Houghton Mifflin Company, page 1159.]
---Vassil.
--
The truly good code is the obviously correct code.
Kent M Pitman wrote:
> Damien Kick <·····@earthlink.net> writes:
>
>> Kent M Pitman wrote:
>>> Vassil Nikolov <···············@pobox.com> writes:
>>>
>>>> (I have also been thinking of using "molecule" for this same
>>>> purpose, but that may be too pretentious.)
>>> If it comes to going in that direction, you might also consider
>>> "compound", since it's a simpler name with a similar connotation.
>>>
>> What about scalar and composite?
>
> [...]
>
> I can see why you'd want to use these two words, but I will here tell
> you why I would not, to the best I can explain such an inherently
> subjective assessment:
Perhaps I've been too badly polluted by my exposure to Perl, which
already uses the terms scalar and composite. Of course, Larry Wall was
first and foremost a linguist before he took to <insert what="witty
comment about creating a language with all the visual appeal of steel
wool with line noise mixed in"> and so he probably knew that
wikidixitisms are actually real words because
<http://en.wikipedia.org/wiki/Philosophical_Investigations>:
<blockquote>
Wittgenstein's method leads to the common summary of Wittgenstein's
argument in the Investigations: "Meaning just is use" � that is, words
are not defined by reference to the objects or things which they
designate in the external world nor by the thoughts, ideas, or mental
representations that one might associate with them, but rather by how
they are used in effective, ordinary communication.
</blockquote>
So he probably knew that enough misuse would eventually worm its way
into use, with the added bonus of giving William Safire heartburn at the
same time.
> Some people consider scalar to refer to numbers, and not necessarily
> other non-subdividable things, such as characters or symbols. (The
> word "atom" should have been the right term, other than due to the
> historical accident of not annexing arrays, etc. as non-atoms.) I
> like to avoid words that have obvious "snags" in them which are
> likely to lead to serious misunderstandings, and this is strong
> enough that it bugs me. I don't assert that others would necessarily
> agree, though; I have no data on that.
I think it is interesting to consider the word "atom"
<http://dictionary.reference.com/browse/atom>:
1.a. A part or particle considered to be an irreducible constituent of a
specified system.
b. The irreducible, indestructible material unit postulated by ancient
atomism.
c. A unit of matter, the smallest unit of an element, having all the
characteristics of that element and consisting of a dense, central,
positively charged nucleus surrounded by a system of electrons. The
entire structure has an approximate diameter of 10-8 centimeter and
characteristically remains undivided in chemical reactions except for
limited removal, transfer, or exchange of certain electrons.
d. This unit regarded as a source of nuclear energy. See Table at
subatomic particle.
The word "atom" was used in the sense of meanings "c" and "d" because at
some point in time it was thought to be the same as "a" and "b". In the
original sense of the word, subatomic should be an oxymoron. Perhaps it
still is an oxymoron <shrug>. It is also interesting to note how the
word was also used to denote affiliation with certain philosophical
schools of thought, Greek atomists with their mechanistic view of
natural order compared with the notion of Platonic forms.
<blockquote cite =
"http://en.wikipedia.org/wiki/Atomism#Atoms_and_ethics">
Three hundred years later, Lucretius in his epic poem On the Nature of
Things would depict Epicurus as the hero who crushed the monster
Religion through educating the people in what was possible in the atoms
and what was not possible in the atoms
</blockquote>
Everything really is political, at the end of the day, isn't it?
In a certain sense, even those things you gave as examples of
non-subdividable things are also composed of subatomic particles. Even
numbers and characters are represented as a sequence of bits. This is
not only true at the implementation level but is exposed at the lisp qua
lisp level with the likes of CHAR-CODE, LDB, ASH, etc. Even SYMBOLs
have SYMBOL-NAMEs. I realize that SYMBOLs are different things than the
STRINGs which make up their SYMBOL-NAMEs. Of course, as a wise man once
said, "there is no uniquely determined equality function for complex
structures--there are only arbitrary ones." From a certain point of
view, does not
(defmethod equivalent-objects ((x symbol) (y string)
(type (eql 'simple-base-string)))
(equal (string x) y))
make sense?
> [...] When I hear compound, I think of something like Lego building
> blocks that has discrete components snapped together, and that's like
> the thing we're talking about. When I hear composite, I don't think
> of something composed of easily identifiable components but rather
> two non-discrete source materials blurred together [...].
>
> [...] the American Heritage Dictionary of the English Language and
> offers a very nice usage summary, a small portion of which I'll quote
> here:
>
> "A compound constitutes a new and independent entity: The school's
> program is a compound of scholarship and athleticism. A composite has
> components that may retain part of their identities: a musical suite
> that is a composite of operatic themes."
>
> -- The American Heritage(R) Dictionary of the English Language,
> Fourth Edition, Copyright (C) 2006 by Houghton Mifflin Company.
> Published by Houghton Mifflin Company. All rights reserved.
>
> as viewed through http://dictionary.reference.com/browse/mixture
Perhaps I'm missing something but I don't think that your source
supports your view. According to your source, it is a composite which
"has components that may retain part of their identities"; a composite
is not the result of "two non-discrete source materials blurred
together", as you phrased it in your own definition. However, I think
this might be making too much ado about nothing. One finds
<http://dictionary.reference.com/browse/composite> the following definition:
<blockquote>
1. made up of disparate or separate parts or elements; compound: a
composite drawing; a composite philosophy.
</blockquote>
I suppose that there might be certain subcultures, e.g chemists or
metallurgy, which make a clear distinction. Of course, if meaning
really just is use, then:
<blockquote>
I do spend a lot of time on [dictionaries], so you might just say I'm
seeing it through the eyes of what I do for a living, but I'm going to
claim for purposes of this article that what unifies [different meanings
of words] is not the [definitions] themselves, but rather the set of
people who provide them. In essence, I'll suggest that [meaning] is a
social phenomenon, akin to a political party, and that what unifies
[definitions] are the people who are its leaders, and the ways in which
they respond (or fail to respond) to the needs of that community.
</blockquote>
And, by that definition, I suppose composite is right out in the Lisp
community <smile>.
From: Kent M Pitman
Subject: Re: terminology for subcategories of ATOM?
Date:
Message-ID: <ups41pur4.fsf@nhplace.com>
Damien Kick <·····@earthlink.net> writes:
> [... Much fun analysis elided ...]
> Perhaps I'm missing something but I don't think that your source
> supports your view. According to your source, it is a composite which
> "has components that may retain part of their identities"; a composite
> is not the result of "two non-discrete source materials blurred
> together", as you phrased it in your own definition.
Well, the irony here is that when components retain their original
identity, you can say that means they don't yield new structure, they
just pass by each other, like so many strangers passing each other on
the busy streets of New York. So, in an odd way, what identifies a
mixture is the tacit agreement of the population not to mix. Whereas,
in a way, compounds do mix, in the sense that there is legitimate
structural interaction. But I can easily see how you'd be confused,
since the text, literally taken (or even taken how the authors
probably meant it), does support your point of view. It just happens
not to disagree with how I was thinking about things. No matter
whether you talk mixtures or compounds, the component parts are not
destroyed. It just enters a different relationship. So probably it's
just a terrible analogy to appeal to at all since it doesn't help to
see the subtleties.
I did take your point about words meaning little more than how they
are used, though. On a related point, I recall Prof. Bill Martin at
MIT, whose computational linguistics class I had the good fortune to
take, speculating about the question of whether if you removed all the
names from a dictionary, it would mean any less. The words give you
starting points for learning things, but he speculated that the mere
shape of a dictionary might be sufficient to determine word meanings,
and that its shape would be so unique that really there could be no
other way of interpreting it than the correct one. This was in the
context of programming a system called XLMS that was a knowledge
representation system in which we created all kinds of typed links
from words to othe words, creating a pretty tight web of knowledge. I
learned a lot about how elusive meaning is from trying (mostly in
vain) to usefully describe a pair of scissors in that language. What
a painful task. I suppose I should have realized I should just
describe it as a composite of a bunch of atoms...
> [... More fun analysis elided ...]
> <blockquote>
> I do spend a lot of time on [dictionaries], so you might just say I'm
> seeing it through the eyes of what I do for a living, but I'm going to
> claim for purposes of this article that what unifies [different meanings
> of words] is not the [definitions] themselves, but rather the set of
> people who provide them. In essence, I'll suggest that [meaning] is a
> social phenomenon, akin to a political party, and that what unifies
> [definitions] are the people who are its leaders, and the ways in which
> they respond (or fail to respond) to the needs of that community.
> </blockquote>
>
> And, by that definition, I suppose composite is right out in the Lisp
> community <smile>.
Just parenthetically speaking, it's always interesting to see what
meanings people will try to inject into one's political writings. ;)
(I'll stop short of saying I got a kick out of it... you probably get
that one way too often.)
On 2007-06-12 02:11:27 -0400, Kent M Pitman <······@nhplace.com> said:
> On a related point, I recall Prof. Bill Martin at
> MIT, whose computational linguistics class I had the good fortune to
> take, speculating about the question of whether if you removed all the
> names from a dictionary, it would mean any less. The words give you
> starting points for learning things, but he speculated that the mere
> shape of a dictionary might be sufficient to determine word meanings,
> and that its shape would be so unique that really there could be no
> other way of interpreting it than the correct one.
reminds me of the zen koan "what is the first word in the dictionary?"
btw, as regards atom/scalar/indivisible/singular v. composite/plural
what would complex numbers be?
From: Christopher C. Stacy
Subject: Re: terminology for subcategories of ATOM?
Date:
Message-ID: <yzlwsy9sl9s.fsf@news.dtpq.com>
Raffael Cavallaro <················@pas-d'espam-s'il-vous-plait-mac.com> writes:
> On 2007-06-12 02:11:27 -0400, Kent M Pitman <······@nhplace.com> said:
>
>> On a related point, I recall Prof. Bill Martin at
>> MIT, whose computational linguistics class I had the good fortune to
>> take, speculating about the question of whether if you removed all the
>> names from a dictionary, it would mean any less. The words give you
>> starting points for learning things, but he speculated that the mere
>> shape of a dictionary might be sufficient to determine word meanings,
>> and that its shape would be so unique that really there could be no
>> other way of interpreting it than the correct one.
>
> reminds me of the zen koan "what is the first word in the dictionary?"
>
> btw, as regards atom/scalar/indivisible/singular v. composite/plural
> what would complex numbers be?
complexities