n-1st? or n-1th?
Joseph Turian
n-1st widget
or the
n-1th widget
?
The former implies that the text is read "n minus first widget".
The latter implies that the text is read "n minus oneth widget".
Also the former sounds better, the second strikes me as more accurate.
Your thoughts?
Joseph
Peter Moylan
It's a question of personal taste, but I'd always say "n minus oneth".
I'd also use parentheses in writing it: (n-1)th.
Indeed, those two observations are related. The other way of saying
it sounds to me as if it should mean "n-(first)".
--
Peter Moylan peter at ee dot newcastle dot edu dot au
http://eepjm.newcastle.edu.au (OS/2 and eCS information and software)
lightbulb
"Joseph Turian" <tur...@gmail.com> wrote in message
news:1110250040.0...@g14g2000cwa.googlegroups.com...
At first I thought you were talking about coordinates. Apparently you're
talking algebra? Could you not just say "widget n minus one", "the widget
at n minus one", or "the n minus one widget?" I think "n minus wunth"
sounds terrible, as does "n minus thirty-thirdth." "n minus forty-fifthth"
is positively lisping.
Mike G.
Don A. Gilmore
"Joseph Turian" <tur...@gmail.com> wrote in message
news:1110250040.0...@g14g2000cwa.googlegroups.com...
"Penultimate"
Don
Kansas City
don groves
Gilmore at eroml...@kc.rr.com hath writ:
For the moment...
--
dg (domain=ccwebster)
Gerald Smyth
I'd say in the mathematical treament of widgets there's a need for the
term itself, that a writearound such as you suggest would come to seem
awkward with repeated use. I prefer 'n minus oneth' for the same reason
I prefer 'nth' - i.e., it is really (n-minus-one)th. Also, those other
examples would be 'n minus thirty-threeth' (=(n-minus-thirty-three)th)
and 'n minus forty-fiveth' (=(n minus forty-five)th)....g
Joseph Turian
lightbulb wrote:
> At first I thought you were talking about coordinates. Apparently
you're
> talking algebra? Could you not just say "widget n minus one", "the
widget
> at n minus one", or "the n minus one widget?" I think "n minus
wunth"
> sounds terrible, as does "n minus thirty-thirdth." "n minus
forty-fifthth"
> is positively lisping.
I agree, I read it again and simply cannot stand it.
So I figured out how to simplify the notation.
(NB _foo_ means italics.)
Original:
A _span_ is a range over (contiguous) words in the input sentence. A
span from the _a_th word through the _b_th word can be given as a pair
_(a, b+1)_, where _a<b+1_. Following the convention in the literature,
_a_ denotes the space between the _a-1_th and _a_th words, and _b+1_
denotes the space between the _b_th and _b+1_th words.
Revised:
A _span_ is a range over (contiguous) words in the input sentence. A
span from word _a_ through word _b_ can be given as pair _(a, b)_,
where _a <= b_.
(screw the convention in the literature.)
JOSEPH
Michael Hamm
> If we were talking about some widget at location n-1, would this be the
> n-1st widget
> or the
> n-1th widget
> ?
>
> The former implies that the text is read "n minus first widget".
> The latter implies that the text is read "n minus oneth widget".
Although the second seems more correct to me, I can report that I've
heard[1] the first more often from mathematicians.
And I do so hereby.
[1] Note "heard". I'm not sure which I've seen more in print, or even on
blackboards.
Michael Hamm
AM, Math, Wash. U. St. Louis
msh...@math.wustl.edu Fine print:
http://math.wustl.edu/~msh210/ ... legal.html
Seán O'Leathlóbhair
This is a tricky one. I think that I would normally use the latter but
it is hard to say it without deliberate thought. I cannot say which I
hear most often, a mixture anyway.
A couple of points.
Several posters seem to think that the latter is more correct but what
if n is 20, 30 etc? I mean, consider the ordinal for 21, it is
twenty-first and not twenty-oneth. Also thirty-first,
hundred-and-first etc. So maybe the former form is more correct (if
that means anything).
On the case of n-33 and other large numbers, I think that it is pretty
much irrelevant. It would be very rare to want to talk about the
(n-33)rd. n - 1 and n + 1 are common, 2 and 3 are less common and
larger numbers even less common. I would not worry about them.
I would say, pick the one that you think sounds best. If possible be
consistent.
--
Seán O'Leathlóbhair
R H Draney
>
>A couple of points.
>
>Several posters seem to think that the latter is more correct but what
>if n is 20, 30 etc? I mean, consider the ordinal for 21, it is
>twenty-first and not twenty-oneth. Also thirty-first,
>hundred-and-first etc. So maybe the former form is more correct (if
>that means anything).
I remember a program in a textbook once that used the "mod" function to apply an
ordinal suffix to a number...a bug had its output talking about the "11st" and
"12nd" in a series....
>On the case of n-33 and other large numbers, I think that it is pretty
>much irrelevant. It would be very rare to want to talk about the
>(n-33)rd. n - 1 and n + 1 are common, 2 and 3 are less common and
>larger numbers even less common. I would not worry about them.
Nobody's mentioned "nth plus one"...it's fraught with ambiguity, but I've heard
it used....
How about parametric offsets?...how do you talk about the n+ith row of a matrix,
for instance?...
>I would say, pick the one that you think sounds best. If possible be
>consistent. =20
Yes...twenty *is* consistent....r
Seán O'Leathlóbhair
R H Draney wrote:
> =?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?= filted:
> >
> >A couple of points.
> >
> >Several posters seem to think that the latter is more correct but
what
> >if n is 20, 30 etc? I mean, consider the ordinal for 21, it is
> >twenty-first and not twenty-oneth. Also thirty-first,
> >hundred-and-first etc. So maybe the former form is more correct (if
> >that means anything).
>
> I remember a program in a textbook once that used the "mod" function
to apply an
> ordinal suffix to a number...a bug had its output talking about the
"11st" and
> "12nd" in a series....
I wrote a numbers to words routine once, to avoid this sort of problem
it took the numbers up to 19 from an array. Only for 20 onwards did I
start to use mod. I have never had cause to write a routine that
outputs ordinal numbers.
> >On the case of n-33 and other large numbers, I think that it is
pretty
> >much irrelevant. It would be very rare to want to talk about the
> >(n-33)rd. n - 1 and n + 1 are common, 2 and 3 are less common and
> >larger numbers even less common. I would not worry about them.
>
> Nobody's mentioned "nth plus one"...it's fraught with ambiguity, but
I've heard
> it used....
Sounds familiar, I may have heard it as well but I don't think that I
would use or recommend it.
> How about parametric offsets?...how do you talk about the n+ith row
of a matrix,
> for instance?...
Ugh.
> >I would say, pick the one that you think sounds best. If possible
be
> >consistent. =20
>
> Yes...twenty *is* consistent....r
I don't know where that odd 20 came from. It was not meant to be
there. (And yes, I know that 20 is even rather than odd.)
Non-mathematicians are often surprised to learn that mathematics
terminology and notation is not standardised. There is not even
universal agreement over whether 0 is positive, negative, neither or
both. When discussing maths, it is usually necessary to start by
agreeing on the meaning of terms and notation. Variation of the sort
being discussed here is minor compared to some of the other problems.
--
Seán O'Leathlóbhair
Ray Heindl
> How about parametric offsets?...how do you talk about the n+ith
> row of a matrix, for instance?...
I generally don't, but if circumstances called for it, I'd say "row
n+i".
N-1th has the advantage of giving us a word that rhymes with "month".
Now we need to figure out a way to work in "n+gry".
--
Ray Heindl
(remove the Xs to reply)
lightbulb
I'm modifying my previously stated opinion. If you're going to be doing
a lot of this sort of notation, you may want to stick to the "(n-1)th"
style, especially if there is a chart or graph which uses "(n-1)." If it
only pops up in one sentence, I'd reword it. Rewording an entire paper
around an awkward, albeit correct, phrasing, can be ineffectual and result
in a paper that is just as awkward as the original. Your discomfort, and
the resulting editing, will show through and bring attention to your writing
instead of your topic.
Mike G.
Aaron Davies
> R H Draney <dado...@spamcop.net> wrote:
>
> > Seán O'Leathlóbhair <jwla...@yahoo.com> wrote:
> >
> > How about parametric offsets?...how do you talk about the n+ith row of a
> > matrix, for instance?...
>
> Ugh.
Seems fairly simple to me: "en plus eyeth". The cell in question occurs
in the "em plus jayth" column, of course.
> > > I would say, pick the one that you think sounds best. If possible
> > > be consistent. =20
> >
> > Yes...twenty *is* consistent....r
>
> I don't know where that odd 20 came from. It was not meant to be there.
> (And yes, I know that 20 is even rather than odd.)
I think Mr. Draney must be using a non-MIME-capable newsreader. An
equals sign followed by two hexadecimal digits (i.e. 0-9 or A-F) is a
quoted-printable escape for the character with that number, in the
charset in use. In particular, 0x20 is 32, the space character. Seán's
post had two trailing spaces after the last sentence, and IIRC, the last
trailing space at the end of a line always has to be escaped, because
otherwise it would be dropped.
--
Aaron Davies
Opinions expressed are solely those of a random number generator.
"I don't know if it's real or not but it is a myth."
-Jami JoAnne of alt.folklore.urban, showing her grasp on reality.
Joseph Turian
Seán O'Leathlóbhair wrote:
> Non-mathematicians are often surprised to learn that mathematics
> terminology and notation is not standardised.
Should there be only one way to say a particular thing?
Someone (Wittgenstein?) said that a good notation can be a more
powerful teacher than equivalent prose.
I didn't understand this until yesterday, when I switched all my
function names from words to Greek symbols. By removing as many names
as possible, I made the problem abstract and could understand it more
clearly. I then realized that, by switching the subscripts with the
function parameters, the reader would have more insight in the nature
of the problem.
With something so rich as communication, why should there be only one
style?
Joseph
Seán O'Leathlóbhair
> Seán O'Leathlóbhair wrote:
> > Non-mathematicians are often surprised to learn that mathematics
> > terminology and notation is not standardised.
>
> Should there be only one way to say a particular thing?
In general no but in particular cases yes. Literature and poetry would
suffer if there was only one way to say anything. On the other hand,
disasters can and do occur because of misinterpretations of technical
information.
These days many people encounter computing before serious mathematics.
In computing there are attempts to standardise things (frequently
unsuccessful) so someone from a computing background is often surprised
when they come to maths and find the lack of standards.
Nonetheless, I don't argue for standard fixed terminology in maths.
The subject is huge and changing. There are not enough words to give
simple fixed names to everything. Names have to be reused and hence it
is usually necessary to indicate their meaning before use. For
example, "field" means at least two utterly different things in maths
and neither involves grass.
> Someone (Wittgenstein?) said that a good notation can be a more
> powerful teacher than equivalent prose.
>
> I didn't understand this until yesterday, when I switched all my
> function names from words to Greek symbols. By removing as many names
> as possible, I made the problem abstract and could understand it more
> clearly. I then realized that, by switching the subscripts with the
> function parameters, the reader would have more insight in the nature
> of the problem.
Once you are used to it, abstraction can indeed be a powerful tool. A
good example is group theory which not only allows a simpler view of
many problems, it provides theorems that can be applied to a very wide
range of problems.
But, it does not work for everyone. Have you not met people who
struggle with simple bits of arithmetic yet do notice if they are short
changed in a shop? They clearly can do basic arithmetic but
abstraction hinders rather than helps them.
> With something so rich as communication, why should there be only one
> style?
> Joseph
--
Seán O'Leathlóbhair
Iskandar Baharuddin
"Seán O'Leathlóbhair" <jwla...@yahoo.com> wrote in message
news:1110360778....@o13g2000cwo.googlegroups.com...
If we seek to achieve uniformity in everything, how do we ensure that
everything is uniformly good?
There is a need for competition in all fields of human endeavour.
Izzy
rban...@shaw.ca
[-]
>If we seek to achieve uniformity in everything, how do we ensure that
>everything is uniformly good?
>
>There is a need for competition in all fields of human endeavour.
>
Trite here, gets you a jail term in other places. Look at that Yukos
guy.
>Izzy
>
Iskandar Baharuddin
rban...@shaw.ca
Gladly. Where are you off to this time?
Joseph Turian
Seán O'Leathlóbhair wrote:
> In general no but in particular cases yes. Literature and poetry
would
> suffer if there was only one way to say anything. On the other hand,
> disasters can and do occur because of misinterpretations of technical
> information.
In *any* domain where clarity is of central importance, the author must
clearly explain all conventions.
> These days many people encounter computing before serious
mathematics.
> In computing there are attempts to standardise things (frequently
> unsuccessful) so someone from a computing background is often
surprised
> when they come to maths and find the lack of standards.
Huh? Math is practiced by narrow cabal of regimented formalists.
I would argue that there are fewer standards in computer science.
That's why there's more buggy software in the world than unsound
proofs.
> Once you are used to it, abstraction can indeed be a powerful tool.
A
> good example is group theory which not only allows a simpler view of
> many problems, it provides theorems that can be applied to a very
wide
> range of problems.
Yes, converting the formulae from an algorithmic flavor to a more
mathematical style has given me a lot of insight into my problem. In
particular, converting long names to single characters as names (e.g.
"Confidence" -> "\chi") has been particularly revealing. I was quite
surprised how much was gained by laying the formulae bare. By stripping
away as much specificity as possible, it's easier to avoid an emotional
attachment that unnecessarily privileges your current solution.
> But, it does not work for everyone. Have you not met people who
> struggle with simple bits of arithmetic yet do notice if they are
short
> changed in a shop? They clearly can do basic arithmetic but
> abstraction hinders rather than helps them.
Huh? I didn't think math could get any more concrete than that
motivated by pecuniary concerns.
Joseph
Mark Brader
| If we were talking about some widget at location n-1, would this be the
| n-1st widget
| or the
| n-1th widget
| ?
Definitely "st" for me -- "n minus first", just as it's 1,341st and
not 1,341th.
Ray Heindl (later):
> N-1th has the advantage of giving us a word that rhymes with "month".
Only in dialects where "one" is pronounced with a U sound rather than
an O sound. Or those where "month" *is* pronounced with an O sound,
I suppose... :-)
--
Mark Brader | "I do have an idea ... based on the quite obvious fact
Toronto | that the number two is ridiculous and can't exist."
m...@vex.net | -- Ben Denison (Isaac Asimov, "The Gods Themselves")
Seán O'Leathlóbhair
Joseph Turian wrote:
> Seán O'Leathlóbhair wrote:
> > In general no but in particular cases yes. Literature and poetry
> would
> > suffer if there was only one way to say anything. On the other
hand,
> > disasters can and do occur because of misinterpretations of
technical
> > information.
>
> In *any* domain where clarity is of central importance, the author
must
> clearly explain all conventions.
Or refer to a definition elsewhere.
> > These days many people encounter computing before serious
> mathematics.
> > In computing there are attempts to standardise things (frequently
> > unsuccessful) so someone from a computing background is often
> surprised
> > when they come to maths and find the lack of standards.
>
> Huh? Math is practiced by narrow cabal of regimented formalists.
> I would argue that there are fewer standards in computer science.
> That's why there's more buggy software in the world than unsound
> proofs.
I am not claiming that standards in computing are very effective (note
my comment in parentheses) but they are attempted. Most people agree
that standards are desirable, they just have trouble agreeing what the
standards should be or following them once they are agreed.
But I don't know a single ISO standard, or anything similar, that
applies to pure mathematics. Nor do I recall serious mathematicians
calling for standards although amateurs sometimes do. It is very
common to hear a mathematician start a presentation or paper by saying:
"By ring I mean . . ." or by referring to another well known
mathematician.
The closest that pure mathematics comes to a standard is that many
people like the terminology or notation of one person and copy him.
The most standard bit of pure maths that I can think of is the use of
the lower case Greek letter pi. I have never seen anything else used
for the famous constant which begin 3.14159. But even this is only one
way, 3.14159... is always written as pi but pi does not always mean
this. It is sometimes used for other things (e.g. a permutation).
As I said elsewhere in the thread, there is not even universal
agreement on whether 0 is positive, negative, neither or both.
> > Once you are used to it, abstraction can indeed be a powerful tool.
> A
> > good example is group theory which not only allows a simpler view
of
> > many problems, it provides theorems that can be applied to a very
> wide
> > range of problems.
>
> Yes, converting the formulae from an algorithmic flavor to a more
> mathematical style has given me a lot of insight into my problem. In
> particular, converting long names to single characters as names (e.g.
> "Confidence" -> "\chi") has been particularly revealing. I was quite
> surprised how much was gained by laying the formulae bare. By
stripping
> away as much specificity as possible, it's easier to avoid an
emotional
> attachment that unnecessarily privileges your current solution.
>
> > But, it does not work for everyone. Have you not met people who
> > struggle with simple bits of arithmetic yet do notice if they are
> short
> > changed in a shop? They clearly can do basic arithmetic but
> > abstraction hinders rather than helps them.
>
> Huh? I didn't think math could get any more concrete than that
> motivated by pecuniary concerns.
Yes, I picked money since it was the most concrete application of
arithmetic that I could think of. My point was that although, you and
I appear find abstraction helpful, there are many who do not. There
are people who can perform arithmetic with money but struggle with the
most trivial bits of abstract algebra.
Evan Kirshenbaum
> The most standard bit of pure maths that I can think of is the use
> of the lower case Greek letter pi. I have never seen anything else
> used for the famous constant which begin 3.14159.
According to Cajori, J. Christoph Sturm, in 1689 used "e". Pi was
first used, by William Oughtred in 1652, to stand for "periphery", and
the constant was "pi/delta" ("diameter"). The use of pi by itself for
the constant was introduced by William Jones in 1706. Euler used "p"
up until about 1736. Bernoulli used "c" (for "circumferentia") before
1740.
So you're just not reading old enough stuff.
--
Evan Kirshenbaum +------------------------------------
HP Laboratories |I value writers such as Fiske.
1501 Page Mill Road, 1U, MS 1141 |They serve as valuable object
Palo Alto, CA 94304 |lessons by showing that the most
|punctilious compliance with the
kirsh...@hpl.hp.com |rules of usage has so little to do
(650)857-7572 |with either writing or thinking
|well.
http://www.kirshenbaum.net/ | --Richard Hershberger
Ray Heindl
> Ray Heindl (later):
>> N-1th has the advantage of giving us a word that rhymes with
>> "month".
>
> Only in dialects where "one" is pronounced with a U sound rather
> than an O sound. Or those where "month" *is* pronounced with an O
> sound, I suppose... :-)
That was the "us" I was referring to: those of us to whom "oneth" and
"month" rhyme.
Michael Hamm
gosled:
> As I said elsewhere in the thread, there is not even universal
> agreement on whether 0 is positive, negative, neither or both.
Can you show me a peer-reviewed paper in English, published within the
past, say, century, that refers to zero as positive, negative, or both?
Perhaps a book will satisfy me instead: it depends on the book.
Seán O'Leathlóbhair
So it seems, I thought that I had some old books but clearly not old
enough. I guess that if I had thought about it more, I would have
realised that it probably took some time before the usage was
standardised.
--
Seán O'Leathlóbhair
Joseph Turian
Michael Hamm wrote:
> Can you show me a peer-reviewed paper in English, published within
the
> past, say, century, that refers to zero as positive, negative, or
both?
Three separate papers, one for each convention, would satisfactorily
support his assertion.
People don't necessarily need to agree that there's disagreement as a
prerequisite for showing that there's disagreement. Otherwise, we'd
never get anywhere.
Joseph
Seán O'Leathlóbhair
I did not say that there was no agreement among English speakers. Here
in the UK, I have never seen zero described as positive or negative. I
believe that the US is equally consistent. The usage is sometimes
termed "Anglo-Saxon" as a result.
I had Bourbaki in mind. I cannot find a good example quickly but a
quick search of the groups suggests that in several European countries,
zero is still considered both positive and negative. Try searching
sci.math for bourbaki and zero, you will find some discussion of the
matter.
This is not the best example of a lack of standards in mathematics. I
selected it because this is not a maths group and it could be easily
understood by non-mathematicians. Here is another example, does a ring
necessarily have an identity? For some authors it does and for others
it does not. I believe that this one does vary among English speakers.
For me, a ring does not necessarily have an identity. I would regard
the even integers as a ring.
Or related to my original example, is zero a member of the natural
numbers?
--
Seán O'Leathlóbhair
Joseph Turian
Seán O'Leathlóbhair wrote:
> Or related to my original example, is zero a member of the natural
> numbers?
Yes, that is a good example. I guess it depends who you ask.
Joseph
Seán O'Leathlóbhair
Which is exactly my point. There are no formal standards in maths.
In some cases, everyone or almost everyone agrees. Leaving aside
Evan's historical examples, pi is in this category. I still would not
call this a standard just a happy coincidence.
In others, there is agreement over a wide area. Zero being neither
positive nor negative is usual in the UK and US.
Sometimes usage varies according to the branch of the subject. An
example occurred in this group some time ago when Evan and I had
similar but not identical interpretations of the mathematical phrase:
"of the order".
And lastly, sometimes the variation seems to be just the personal
preference of the author. Whether zero is a natural number of not
seems to be an example of this category. For me, the natural numbers
start at 1 but I don't assume this when I hear someone else talk about
the natural numbers.
--
Seán O'Leathlóbhair
don groves
mh...@artsci.wustl.edu hath writ:
> Yesterday, ?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg <jwla...@yahoo.com>
> gosled:
> > As I said elsewhere in the thread, there is not even universal
> > agreement on whether 0 is positive, negative, neither or both.
>
> Can you show me a peer-reviewed paper in English, published within the
> past, say, century, that refers to zero as positive, negative, or both?
>
> Perhaps a book will satisfy me instead: it depends on the book.
Only time I've ever heard of signed zero was back in the days of
ones-compliment computer aritmetic when we had zero and minus
zero.
--
dg (domain=ccwebster)
John Dawkins
mh...@artsci.wustl.edu (Michael Hamm) wrote:
> Yesterday, ?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg <jwla...@yahoo.com>
> gosled:
> > As I said elsewhere in the thread, there is not even universal
> > agreement on whether 0 is positive, negative, neither or both.
>
> Can you show me a peer-reviewed paper in English, published within the
> past, say, century, that refers to zero as positive, negative, or both?
In Bourbakish, "positive" means "greater than or equal to 0" (as opposed
to "strictly positive", which means ">0") and "negative" means "less
than or equal to 0", so 0 is both positive and negative.
--
J.
Aaron Davies
I believe IEEE floating point math has negative zero as well. I think
it's the result of division by negative infinity.
don groves
central.com.invalid>, Aaron Davies at aa...@avalon.pascal-
central.com.invalid hath writ:
> don groves <dgr...@domain.net(seesigfordomain)> wrote:
>
> > In article <d0rium$8r2$1...@newsreader.wustl.edu>, Michael Hamm at
> > mh...@artsci.wustl.edu hath writ:
> > > Yesterday, ?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg <jwla...@yahoo.com>
> > > gosled:
> > > > As I said elsewhere in the thread, there is not even universal
> > > > agreement on whether 0 is positive, negative, neither or both.
> > >
> > > Can you show me a peer-reviewed paper in English, published within the
> > > past, say, century, that refers to zero as positive, negative, or both?
> > >
> > > Perhaps a book will satisfy me instead: it depends on the book.
> >
> > Only time I've ever heard of signed zero was back in the days of
> > ones-compliment computer aritmetic when we had zero and minus
> > zero.
>
> I believe IEEE floating point math has negative zero as well. I think
> it's the result of division by negative infinity.
it myself.
--
dg (domain=ccwebster)
Seán O'Leathlóbhair
Computers and maths are quite different things. They frequently use
the same words, and occasionally, they mean the same thing by them.
But very often they use them differently.
I don't like to see maths corrupted by ideas from computing. How
computers handle positive, negative and zero has no bearing on maths.
In most computer architectures you could say that zero had a sign.
Usually only one bit is used for the sign (more would be a waste) and
usually it is set one way for positive and the other for negative. So,
since a bit can have only two values, it will be set in one of these
ways for zero. There are two possibilities, it may be set in only one
way for zero. The common integer format 2's complement is an example
and you could say that zero is positive. The other possibility is that
the setting of the sign bit does not matter for zero and you have the
possibility of distinct positive and negative zeros. But none of this
says anything about mathematics. Most English speaking mathematicians
say that zero is neither negative nor positive. Some odd followers of
Bourbaki say that zero is both positive and negative. I am not aware
of any that say that zero is positive but not negative or vice versa.
--
Seán O'Leathlóbhair
rban...@shaw.ca
<jwla...@yahoo.com> wrote:
>
>I don't like to see maths corrupted by ideas from computing. How
>computers handle positive, negative and zero has no bearing on maths.
>In most computer architectures you could say that zero had a sign.
>Usually only one bit is used for the sign (more would be a waste) and
>usually it is set one way for positive and the other for negative. So,
>since a bit can have only two values, it will be set in one of these
>ways for zero. There are two possibilities, it may be set in only one
>way for zero. The common integer format 2's complement is an example
>and you could say that zero is positive. The other possibility is that
>the setting of the sign bit does not matter for zero and you have the
>possibility of distinct positive and negative zeros. But none of this
>says anything about mathematics. Most English speaking mathematicians
>say that zero is neither negative nor positive. Some odd followers of
>Bourbaki say that zero is both positive and negative. I am not aware
>of any that say that zero is positive but not negative or vice versa.
Enter Katy Katy and the demolition team, wot?
John Lawler
>>Michael Hamm at mh...@artsci.wustl.edu writes:
><jwla...@yahoo.com> writes
>> > > As I said elsewhere in the thread, there is not even universal
>> > > agreement on whether 0 is positive, negative, neither or both.
>> > Can you show me a peer-reviewed paper in English, published within
>> > the past, say, century, that refers to zero as positive, negative,
>> > or both?
>> > Perhaps a book will satisfy me instead: it depends on the book.
>> Only time I've ever heard of signed zero was back in the days of
>> ones-complement computer aritmetic when we had zero and minus
>> zero.
>Computers and maths are quite different things. They frequently use
>the same words, and occasionally, they mean the same thing by them.
>But very often they use them differently.
>I don't like to see maths corrupted by ideas from computing. How
>computers handle positive, negative and zero has no bearing on maths.
>In most computer architectures you could say that zero had a sign.
>Usually only one bit is used for the sign (more would be a waste) and
>usually it is set one way for positive and the other for negative. So,
>since a bit can have only two values, it will be set in one of these
>ways for zero. There are two possibilities, it may be set in only one
>way for zero. The common integer format 2's complement is an example
>and you could say that zero is positive. The other possibility is that
>the setting of the sign bit does not matter for zero and you have the
>possibility of distinct positive and negative zeros. But none of this
>says anything about mathematics. Most English speaking mathematicians
>say that zero is neither negative nor positive. Some odd followers of
>Bourbaki say that zero is both positive and negative. I am not aware
>of any that say that zero is positive but not negative or vice versa.
Don't let's forget that it's not an empirical matter whether zero has a sign
or not (except in computing, as you point out). The number zero is a mental
construct, defined by humans, not an objective physical phenomenon.
As to whether it's in the set of Natural Numbers, historically it wasn't,
since it was invented in historic times, long after counting started,
but for convenience' sake that set can be defined as containing it, if
anybody wants to. It's just a set, and it consists of precisely what it's
defined to contain; mathematical objects have a humptidumptian epistemology.
-John Lawler http://www.umich.edu/~jlawler Michigan Linguistics
----------------------------------------------------------------
'God made the Natural Numbers; all the rest is the work of man.'
-- Leopold Kronecker
Seán O'Leathlóbhair
That's what I was trying to say but phrased differently and maybe
better. Here's another version:
If a computer has an instruction called "Is positive?" then we can test
or determine how it responds to the computer's representation of zero.
But mathematical zero is a mental construct and the only way to be
certain whether someone considers zero to be positive is to ask him.
Most English speaking mathematicians alive today would say "No" but
some others would say "Yes".
> As to whether it's in the set of Natural Numbers, historically it
wasn't,
> since it was invented in historic times, long after counting started,
> but for convenience' sake that set can be defined as containing it,
if
> anybody wants to. It's just a set, and it consists of precisely what
it's
> defined to contain; mathematical objects have a humptidumptian
epistemology.
I had not thought of the historical point of view, certainly not back
as far the invention of zero. When I am discussing mathematics, "The
natural numbers" is an idiom whose meaning cannot be accurately deduced
from the everyday meanings of "natural" and "number". You could
consider "The natural numbers" to be a proper name for a particular
set. It would be possible for all living mathematicians to agree
whether zero was a member of this set but unfortunately they do not.
The following exchange is common among mathematicians who do not know
each other well:
"All natural numbers have the property that . . ."
"Do you include zero in the natural numbers?"
Was the phrase "The natural numbers", as used by mathematicians today,
coined before the invention of zero? I, assumed, that it dated from
the formalisation of set theory and the foundations of mathematics
which is much more recent than zero.
It is like "John Lawler". In an ideal world it would uniquely identify
a single human being but the two of us know that it does not. In the
same way, it would be ideal if "The natural numbers" identified a
unique set but again it does not. I am only aware of two common
meanings but others may exist.
> -John Lawler http://www.umich.edu/~jlawler Michigan Linguistics
> ----------------------------------------------------------------
> 'God made the Natural Numbers; all the rest is the work of man.'
> -- Leopold Kronecker
--
Seán O'Leathlóbhair
A non-unique John Lawler
Seán O'Leathlóbhair
Murray Arnow wrote:
> "=?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?=" <jwla...@yahoo.com>
wrote:
> >
> > John Lawler wrote:
> > > Se=E1n O'Leathl=F3bhair <jwla...@yahoo.com> writes:
> >
> > It is like "John Lawler". In an ideal world it would uniquely
identify
> > a single human being but the two of us know that it does not. In
the
> > same way, it would be ideal if "The natural numbers" identified a
> > unique set but again it does not. I am only aware of two common
> > meanings but others may exist.
> >
>
identities
> and answers his own posts? BTW, what is all this ISO stuff in the
From
> header?
We have had this discussion before and not come to any conclusion.
You, the reader, need to drawn your own conclusions.
As to funny characters in the from header, some newsreaders appear to
cope with them and others don't. Here is my name with the accents
omitted: Sean O'Leathlobhair. The correct form has an accent similar
the French acute on the a of Sean and the o of lobhair.
--
Seán O'Leathlóbhair
K. Edgcombe
>Don't let's forget that it's not an empirical matter whether zero has a sign
>or not (except in computing, as you point out). The number zero is a mental
>construct, defined by humans, not an objective physical phenomenon.
>
(This is not really a reply to John Lawler; I've just picked up his posting to
respond to).
But in the argument as whether it is positive, the discussion is surely about
the meaning of "positive", not about the properties (intrinsic or conventional)
of zero. Most English-speaking mathematicians opt for "positive",
"negative", "non-negative" and (if necessary) "non-positive" to provide
an adequate vocabulary for talking about the different classes. They do
these things, I gather, differently in France. But the important mental
construct "zero" is the same in both places; it's "positive" that's different.
As for the natural numbers, I teach both mathematicians and computer scientists
and am constantly having to switch conventions; but there's no point in getting
theological about it. What matters is that whichever set you take to be the
natural numbers, you have a way to talk about the other set (including or
excluding zero, as necessary) when you need to.
Katy
Lars Enderin
> Murray Arnow wrote:
>
>>"=?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?=" <jwla...@yahoo.com>
>
> wrote:
>
>>>John Lawler wrote:
>>>
>>>>Se=E1n O'Leathl=F3bhair <jwla...@yahoo.com> writes:
>>>
>>and answers his own posts? BTW, what is all this ISO stuff in the
>
> From
>
>>header?
>
> As to funny characters in the from header, some newsreaders appear to
> cope with them and others don't. Here is my name with the accents
> omitted: Sean O'Leathlobhair. The correct form has an accent similar
> the French acute on the a of Sean and the o of lobhair.
>
> --
> Seán O'Leathlóbhair
>
Apparently, Murray Arnow's news reader (News Xpress 2.01) is not capable
of decoding your name in the From header, which is iso-8859-1 encoded
according to the MIME standard as "base 64", e g all characters are
treated as binary and transformed to a string of seven-bit characters. A
more readable representation (quoted printable) is seen in (the other)
John Lawler's attribution.
I think you can say that the accent is *identical* to the (not
exclusively French) acute accent.
Seán O'Leathlóbhair
K. Edgcombe wrote:
> >
> >Don't let's forget that it's not an empirical matter whether zero
has a sign
> >or not (except in computing, as you point out). The number zero is
a mental
> >construct, defined by humans, not an objective physical phenomenon.
> >
>
> (This is not really a reply to John Lawler; I've just picked up his
posting to
> respond to).
(Which John Lawler is it not a reply to?)
> But in the argument as whether it is positive, the discussion is
surely about
> the meaning of "positive", not about the properties (intrinsic or
conventional)
> of zero. Most English-speaking mathematicians opt for "positive",
> "negative", "non-negative" and (if necessary) "non-positive" to
provide
> an adequate vocabulary for talking about the different classes. They
do
> these things, I gather, differently in France. But the important
mental
> construct "zero" is the same in both places; it's "positive" that's
different.
A valid point. There is much better agreement on properties such as 0
being the additive identity i.e. x + 0 = x for any x (we should specify
the domain of x).
> As for the natural numbers, I teach both mathematicians and computer
scientists
> and am constantly having to switch conventions; but there's no point
in getting
> theological about it. What matters is that whichever set you take to
be the
> natural numbers, you have a way to talk about the other set
(including or
> excluding zero, as necessary) when you need to.
In my experience, mathematicians differ on the question of whether the
natural numbers include zero.
> Katy
As we have a few times before, we are drifting out of the domain of
this newsgroup. I am quite happy talking maths but I am aware that
others in this group may not find it so interesting. By coincidence
there is a thread entitled "the sign of zero" running in sci.math at
the moment. But it is a mixture of silly stuff and abstruse stuff and
would not help this thread much. If you search the archives of
sci.math, you will find threads like this one. "Bourbaki" is a good
keyword to find relevant discussions.
--
Seán O'Leathlóbhair
Seán O'Leathlóbhair
Do you mean the other John's own name or his reference to me? He uses
the Anglicised form of our shared name and hence has no accents. That
my Irish form can go to other John and back to me or on to you without
distortion does not indicate whether or not Murray ever saw it
correctly.
> I think you can say that the accent is *identical* to the (not
> exclusively French) acute accent.
In visual form yes but in name or function no. Also, (as far as I
know) the French only use this accent on e whereas Irish may use it on
any vowel. The Spanish also use the same accent on any vowel. If a
program can display French correctly, it will probably but not
certainly manage Irish. But, if it can display Spanish correctly, it
should certainly cope with Irish.
--
Seán O'Leathlóbhair
K. Edgcombe
=?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?= <jwla...@yahoo.com> wrote:
>
>K=2E Edgcombe wrote:
>>
>> (This is not really a reply to John Lawler; I've just picked up his
>posting to
>> respond to).
>
>(Which John Lawler is it not a reply to?)
The one who wrote what I quoted, whoever he may be; but I suppose I could
say it wasn't a reply to either (or any) of the John Lawlers.
>As we have a few times before, we are drifting out of the domain of
>this newsgroup. I am quite happy talking maths but I am aware that
>others in this group may not find it so interesting. By coincidence
>there is a thread entitled "the sign of zero" running in sci.math at
>the moment. But it is a mixture of silly stuff and abstruse stuff and
>would not help this thread much. If you search the archives of
>sci.math, you will find threads like this one. "Bourbaki" is a good
>keyword to find relevant discussions. =20
Been there (sci.math), done that.
There's often a surprising amount of response on this group to discussions
which are partly mathematical. (Please will nobody ask whether zero is even?)
I think such discussions are often also partly linguistic; notably so when
participants think they are dealing with a reality and find they are dealing
only with a linguistic convention. This process has many parallels outside
maths, of course.
Katy
Seán O'Leathlóbhair
K. Edgcombe wrote:
> In article <1110821824....@f14g2000cwb.googlegroups.com>,
> =?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?= <jwla...@yahoo.com>
wrote:
> >
> >K=2E Edgcombe wrote:
> >>
> >> (This is not really a reply to John Lawler; I've just picked up
his
> >posting to
> >> respond to).
> >
> >(Which John Lawler is it not a reply to?)
>
> The one who wrote what I quoted, whoever he may be; but I suppose I
could
> say it wasn't a reply to either (or any) of the John Lawlers.
Looking again more carefully I see that it was the other one. So I
won't be offended by being not replied to.
> >As we have a few times before, we are drifting out of the domain of
> >this newsgroup. I am quite happy talking maths but I am aware that
> >others in this group may not find it so interesting. By coincidence
> >there is a thread entitled "the sign of zero" running in sci.math at
> >the moment. But it is a mixture of silly stuff and abstruse stuff
and
> >would not help this thread much. If you search the archives of
> >sci.math, you will find threads like this one. "Bourbaki" is a good
> >keyword to find relevant discussions. =20
>
> Been there (sci.math), done that.
>
> There's often a surprising amount of response on this group to
discussions
> which are partly mathematical. (Please will nobody ask whether zero
is even?)
> I think such discussions are often also partly linguistic; notably so
when
> participants think they are dealing with a reality and find they are
dealing
> only with a linguistic convention. This process has many parallels
outside
> maths, of course.
Yes, we have drifted in maths a number of times. Usually it starts as
a fairly topical linguistic question with a mathematical flavour and
then drifts deeper and deeper in mathematics. This one has retained
some valid linguistic elements but the "exponential growth" thread went
quite deep into maths and I must accept a large portion of the
responsibility. I am trying to avoid doing it too often.
I promise not to discuss whether zero is even, at least not in this
group.
One of my personal crusades is to discourage the corruption of
mathematical language by that of computers. As I said a few posts ago,
they share many words and sometimes mean the same thing by them but
very often they don't mean the same thing in the two disciplines.
It happens in other disciplines as well. People learn a specialised
meaning of a word in some discipline and expect others, including the
general public, to share this specialised meaning. A friend studying
psychology used to do this often.
> Katy
--
Seán O'Leathlóbhair
Don Aitken
wrote:
>Lars Enderin <lars.e...@gmail.com> wrote:
>> Seán O'Leathlóbhair wrote:
>> > Murray Arnow wrote:
>> >
>> >>"=?iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg==?=" <jwla...@yahoo.com>
>> > wrote:
>> >
>> >>>John Lawler wrote:
>> >>>
>> >>>>Se=E1n O'Leathl=F3bhair <jwla...@yahoo.com> writes:
>>
>> >>and answers his own posts? BTW, what is all this ISO stuff in the
>> >>From header?
>> >
>> > As to funny characters in the from header, some newsreaders appear to
>> > cope with them and others don't. Here is my name with the accents
>> > omitted: Sean O'Leathlobhair. The correct form has an accent similar
>> > the French acute on the a of Sean and the o of lobhair.
>> >
>> Apparently, Murray Arnow's news reader (News Xpress 2.01) is not capable
>> of decoding your name in the From header, which is iso-8859-1 encoded
>> according to the MIME standard as "base 64", e g all characters are
>> treated as binary and transformed to a string of seven-bit characters. A
>> more readable representation (quoted printable) is seen in (the other)
>> John Lawler's attribution.
>> I think you can say that the accent is *identical* to the (not
>> exclusively French) acute accent.
>
>It's more than apparent; it's the goods. Isn't there a standard for what
>is allowable in news post headers?
I believe that the relevant RFCs specify standard us-ascii text, so no
accented characters at all. In practice, they are frequent enough,
especially in groups where the language is not English, that there is
a de-facto standard which most newsreaders can cope with. Yours is the
only one I have noticed for which it is a problem. The definitive
USEFOR standard for all such things has been due real soon now for
about ten years, and sensible software writers have stopped holding
their breath.
--
Don Aitken
Mail to the addresses given in the headers is no longer being
read. To mail me, substitute "clara.co.uk" for "freeuk.com".
Michael Hamm
> > > Apparently, Murray Arnow's news reader (News Xpress 2.01) is not
> >
> > It's more than apparent; it's the goods. Isn't there a standard for
> > what is allowable in news post headers?
>
> I believe that the relevant RFCs specify standard us-ascii text, so no
> accented characters at all. In practice, they are frequent enough,
> especially in groups where the language is not English, that there is
> a de-facto standard which most newsreaders can cope with. Yours is the
> only one I have noticed for which it is a problem.
Mine's another, fwiw. See headers for client and version number.
Michael Hamm
gosled:
> As we have a few times before, we are drifting out of the domain of
> this newsgroup. I am quite happy talking maths but I am aware that
> others in this group may not find it so interesting.
Whyever?
Don Aitken
(Michael Hamm) wrote:
>Today, Don Aitken <don-a...@freeuk.com> gosled:
>> > > Apparently, Murray Arnow's news reader (News Xpress 2.01) is not
>> > > capable of decoding your name in the From header <snip>
>> >
>> > It's more than apparent; it's the goods. Isn't there a standard for
>> > what is allowable in news post headers?
>>
>> I believe that the relevant RFCs specify standard us-ascii text, so no
>> accented characters at all. In practice, they are frequent enough,
>> especially in groups where the language is not English, that there is
>> a de-facto standard which most newsreaders can cope with. Yours is the
>> only one I have noticed for which it is a problem.
>
>Mine's another, fwiw. See headers for client and version number.
>
Since tin has been around for a pretty long time, I suspect that
ver.1.2 might be getting a bit long in the tooth. www.tin.org offers
versions up to 1.6.2 and 1.7.7. It probably says somewhere on that
site if the problem has been fixed.
Peter Moylan
>It happens in other disciplines as well. People learn a specialised
>meaning of a word in some discipline and expect others, including the
>general public, to share this specialised meaning. A friend studying
>psychology used to do this often.
I met an interesting case of this a couple of years ago. A student
at another university wrote to me - having tracked down my name via
my publications, I imagine - to get my definition of "positive
feedback". Apparently a computer science lecturer there was giving
a course on control and stability, and was using the term "positive
feedback" to mean what everyone else in the systems theory community
calls "negative feedback". By pure coincidence I happened to know
the lecturer involved. I didn't have the courage to tell her that she
was teaching rubbish, but it did mean that I could guess the reason
for the confusion.
In control and systems theory, whose practitioners are mostly
engineers, sometimes mathematicians, and occasionally from other
places such as economics and biology, the terms have a well-understood
meaning. Stripped of the technical jargon, the approximate meanings
are:
- Negative feedback is where some part of the output of a process
is subtracted from the input. The goal of this is, crudely
speaking, to correct any changes in the output by reacting to
its wandering away from where it should be and pushing it back.
- Positive feedback is where a part of the output is _added_ to
the input. The effect of this is to amplify any change in the
output.
Occasionally positive feedback is useful, but most of the time what
you want is negative feedback. Negative feedback stabilises a system
and gives it various other desirable properties. Positive feedback
amplifies any imperfections in a system, and can destabilise it.
In other words, negative feedback is a Good Thing, and positive
feedback is a Bad Thing.
In the social sciences, however, positive feedback is seen as a
Good Thing because it helps people's self-esteem; and negative
feedback is a euphemism for criticism.
In the example that came to my notice, the lecturer involved had
come to computer science via a business background, and therefore
probably had more of a touchy-feely approach to technology than the
typical computer scientist has. Ergo, confusion over terminology.
--
Peter Moylan peter at ee dot newcastle dot edu dot au
http://eepjm.newcastle.edu.au (OS/2 and eCS information and software)
Martin Ambuhl
> In the social sciences, however, positive feedback is seen as a
> Good Thing because it helps people's self-esteem; and negative
> feedback is a euphemism for criticism.
This is overbroad, at the very least. The social science I deal with as
a political economist are political science and economics. The people I
deal with are well-versed in systems theory and do not use "positive
feedback" as a touchy-feely Good Thing.
I note that the _Oxford Dictionary of Psychology_ also defines negative
and positive feedback in the systems-theoretic way, as does even the
_Cambridge Dictionart of Philosophy_.
I know of no reason for your assertion. What social sciences did you
have in mind? Please remember that journalism, business, and
Derrida-infected literature majors are not social scientists.
R J Valentine
Works both ways, though. Many, if not most, computer people get "simplex"
and "half-duplex" switched (as far as I can tell because an ANSI or ISO
glossary was based almost to the word on an IBM glossary (which Truly
[say, ...] had nothing to do with) that had it backwards. Now there's no
convincing people, because they think they understand it (though emes like
"simple" and "half" have to be ignored to "understand it").
--
R. J. Valentine <mailto:r...@smart.net>
IANASSBIHADISS
Evan Kirshenbaum
> In control and systems theory, whose practitioners are mostly
> engineers, sometimes mathematicians, and occasionally from other
> places such as economics and biology, the terms have a well-understood
> meaning. Stripped of the technical jargon, the approximate meanings
> are:
> - Negative feedback is where some part of the output of a process
> is subtracted from the input. The goal of this is, crudely
> speaking, to correct any changes in the output by reacting to
> its wandering away from where it should be and pushing it back.
> - Positive feedback is where a part of the output is _added_ to
> the input. The effect of this is to amplify any change in the
> output.
>
> Occasionally positive feedback is useful, but most of the time what
> you want is negative feedback. Negative feedback stabilises a system
> and gives it various other desirable properties. Positive feedback
> amplifies any imperfections in a system, and can destabilise it.
> In other words, negative feedback is a Good Thing, and positive
> feedback is a Bad Thing.
I suspect that in many domains (including my own), "feedback" is
confounded with the somewhat similar notion of "reinforcement", in
which parameters to the system are modified (directly or indirectly)
in order to make the system behave more or less like it is currently
behaving.
In a learning system (whether biological or computational), positive
reinforcement is often important early on to get the system doing
basically the right thing ("do like you did, but more so") and then
negative reinforcement becomes more important to fine tune things so
it's less likely to do the wrong thing (or go "too far").
What makes it a bit confusing is that the signal used to determine the
magnitude and direction of the reinforcement based on observation of
the behavior is often called a "feedback" signal. So "feedback that
indicates that the system should be positively reinforced" becomes
"positive feedback".
They're similar notions, but not identical, though on reflection they
may be less different than I implied.
--
Evan Kirshenbaum +------------------------------------
HP Laboratories |Bullwinkle: You sure that's the
1501 Page Mill Road, 1U, MS 1141 | only way?
Palo Alto, CA 94304 |Rocky: Well, if you're going to be
| a hero, you've got to do
kirsh...@hpl.hp.com | stupid things every once in
(650)857-7572 | a while.
Aaron Davies
Psychology, I imagine. I'd also include management, but you've ruled out
business, so I won't.
Peter Moylan
>Today, iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg <jwla...@yahoo.com>
>gosled:
>> As we have a few times before, we are drifting out of the domain of
>> this newsgroup. I am quite happy talking maths but I am aware that
>> others in this group may not find it so interesting.
>
>Whyever?
It's likely that not a few denizens of this group have sampled
sci.math at one time or another, and found it full of undergraduates
with so many misconceptions that it was too hard to separate the
wheat from the chaff.
The charter for alt.usage.english seems to have gone lost, but by
custom the charter has become "discussions about anything", preferably
with people who have a clue. It's that last rider that's the
sticking point in many other newsgroups.
Charles Riggs
<mam...@earthlink.net> wrote:
>Peter Moylan wrote:
>
>> In the social sciences, however, positive feedback is seen as a
>> Good Thing because it helps people's self-esteem; and negative
>> feedback is a euphemism for criticism.
>
>This is overbroad, at the very least. The social science I deal with as
>a political economist are political science and economics. The people I
>deal with are well-versed in systems theory and do not use "positive
>feedback" as a touchy-feely Good Thing.
When would positive feedback in a social setting be a Bad Thing? That
was Peter's point: it is generally a bad thing in electronics, but is
usually seen as being helpful when someone gives someone else positive
feedback.
--
Charles Riggs
There are no accented letters in my email address
Matti Lamprhey
>
> [...] Many, if not most, computer people get "simplex"
> and "half-duplex" switched (as far as I can tell because an ANSI or
> ISO glossary was based almost to the word on an IBM glossary (which
> Truly [say, ...] had nothing to do with) that had it backwards. Now
> there's no convincing people, because they think they understand it
> (though emes like "simple" and "half" have to be ignored to
> "understand it").
I'd be surprised if this confusion is as extensive as you say. Let's
take me as an example -- I think I know what the terms mean, and I'm not
going to check before exposing my potential confusion. Am I right in
thinking that both terms label one-way communication channels,
"half-duplex" being used for those whose direction can be reversed and
"simplex" for those which can't?
And what's an eme?
Matti
Jitze Couperus
wrote:
>
>When would positive feedback in a social setting be a Bad Thing? That
>was Peter's point: it is generally a bad thing in electronics, but is
>usually seen as being helpful when someone gives someone else positive
>feedback.
Positive feedback is not always a good thing.
If you wantr to encourage a certain behaviour, then yes.
But if you want to discourage a certain behaviour, then
negative feedback is called for.
I like the way you are now paying more attention - keep it up!
I am disappointed that you still blather on at great length
on matters you know not of. This really does not improve
your net persona.
Jitze
Seán O'Leathlóbhair
That's a nice example of the phenomenon that I was referring to. The
popularity of eBay has made the terms "positive feedback" and "negative
feedback" well known with "positive" good and "negative" bad. But, as
you say, in other contexts, positive feedback is not a good thing.
Examples are the screech when a microphone is too near to a speaker and
a nuclear explosion. An example of good negative feedback are the
governors that you see on the huge old steam engines in the London
Science Museum.
Negative feedback is sometimes seen as a good thing in a social
context, how about a teacher's criticism of his student's work? How
much would the student learn with no negative feedback to stop him
wandering off course?
--
Seán O'Leathlóbhair
Seán O'Leathlóbhair
Peter Moylan wrote:
> Michael Hamm turpitued:
> >Today, iso-8859-1?B?U2XhbiBPJ0xlYXRobPNiaGFpcg <jwla...@yahoo.com>
> >gosled:
> >> As we have a few times before, we are drifting out of the domain
of
> >> this newsgroup. I am quite happy talking maths but I am aware
that
> >> others in this group may not find it so interesting.
> >
> >Whyever?
>
> It's likely that not a few denizens of this group have sampled
> sci.math at one time or another, and found it full of undergraduates
> with so many misconceptions that it was too hard to separate the
> wheat from the chaff.
Stay there a while and you will get to know the good guys. As well as
a large number of cranks and confused people, there are some very
knowledgeable people. I carefully select what I read based on who
posts it. These knowledgeable people are usually very willing to help
and provided you read their replies carefully and reply sensibly, they
are likely to work hard to help you. As in any news group, it is worth
looking through it first to see if your question is already there.
They will not appreciate yet another thread on whether 0.9999... = 1 or
whether 0 is even.
> The charter for alt.usage.english seems to have gone lost, but by
> custom the charter has become "discussions about anything",
preferably
> with people who have a clue. It's that last rider that's the
> sticking point in many other newsgroups.
Some digressions seem especially popular. Food is common and not too
surprising but the tendency to drift into maths is more surprising.
> --
> Peter Moylan peter at ee dot newcastle dot edu dot au
> http://eepjm.newcastle.edu.au (OS/2 and eCS information and software)
--
Seán O'Leathlóbhair
John Ings
<jwla...@yahoo.com> wrote:
> But, as
>you say, in other contexts, positive feedback is not a good thing.
Without positive feedback to cause oscillation no radio or TV could
work.
>Examples are the screech when a microphone is too near to a speaker and
>a nuclear explosion.
Where's the feedback in a nuclear explosion?
Seán O'Leathlóbhair
John Ings wrote:
> On 23 Mar 2005 01:33:41 -0800, "Seán O'Leathlóbhair"
> <jwla...@yahoo.com> wrote:
>
> > But, as
> >you say, in other contexts, positive feedback is not a good thing.
>
> Without positive feedback to cause oscillation no radio or TV could
> work.
I am not trying to say that positive feedback is always bad either.
Sometimes good, sometimes bad.
However I thought that oscillations were an example of negative
feedback. For example in a pendulum there is an acceleration
(approximately) proportional to the distance from the centre but in the
opposite direction (which I would call negative). If it were in the
same direction (positive), the pendulum would not swing but shoot off
like a rocket. Oscillations in electrical circuits are similar, a
tuned circuit usually involves an inductor (coil) and a capacitor. As
the charge builds up in the capacitor, it works against the current
flow, and like the pendulum, eventually stops and reverses the flow.
If you want some maths, in both cases you get a differential equation
like this: f''(x) = -kf(x) which says that the acceleration (second
derivative) is a negative constant times the current value. The
solution is a sine wave.
> >Examples are the screech when a microphone is too near to a speaker
and
> >a nuclear explosion.
>
> Where's the feedback in a nuclear explosion?
One atom splits and the shards cause a few more to split. The shards
from these few more split a few more each and the reaction grows
exponentially (*) until the plutonium is all used up or blown all over
the place.
(*) This is a mathematically appropriate use of the term.
This time the differential equation is f'(x) = kf(x) which says that
the rate of change is a positive constant times the current value. The
solution is the exponential function.
We are way off topic again.
--
Seán O'Leathlóbhair
John Ings
<jwla...@yahoo.com> wrote:
>> > But, as
>> >you say, in other contexts, positive feedback is not a good thing.
>>
>> Without positive feedback to cause oscillation no radio or TV could
>> work.
>
>I am not trying to say that positive feedback is always bad either.
>Sometimes good, sometimes bad.
>
>However I thought that oscillations were an example of negative
>feedback.
No. Your example of positive feedback from a microphone too near a
loudspeaker is an example. That squeal is oscillation.
>For example in a pendulum there is an acceleration
>(approximately) proportional to the distance from the centre but in the
>opposite direction (which I would call negative). If it were in the
>same direction (positive), the pendulum would not swing but shoot off
>like a rocket.
Acceleration from what? Gravity? That's alternately acceleration and
deceleration and in neither case feedback.
> Oscillations in electrical circuits are similar, a
>tuned circuit usually involves an inductor (coil) and a capacitor. As
>the charge builds up in the capacitor, it works against the current
>flow, and like the pendulum, eventually stops and reverses the flow.
And left alone, dampens down in meek obedience to the 2nd law of
thermodynamics. There's no feedback there either.
Now add an amplifier --transistor or electron tube-- add positive
feedback, place the coil and capacitor in the circuit. Now you have an
tuned oscillator.
>> Where's the feedback in a nuclear explosion?
>
>One atom splits and the shards cause a few more to split. The shards
>from these few more split a few more each and the reaction grows
>exponentially (*) until the plutonium is all used up or blown all over
>the place.
That's not feedback.
>We are way off topic again.
Yeah, but ain't it fun?
Seán O'Leathlóbhair
John Ings wrote:
> On 23 Mar 2005 06:05:57 -0800, "Seán O'Leathlóbhair"
> <jwla...@yahoo.com> wrote:
>
> >> > But, as
> >> >you say, in other contexts, positive feedback is not a good
thing.
> >>
> >> Without positive feedback to cause oscillation no radio or TV
could
> >> work.
> >
> >I am not trying to say that positive feedback is always bad either.
> >Sometimes good, sometimes bad.
> >
> >However I thought that oscillations were an example of negative
> >feedback.
>
> No. Your example of positive feedback from a microphone too near a
> loudspeaker is an example. That squeal is oscillation.
That is actually a bit of both. The oscillation is caused by negative
feedback (as I believe oscillations usually are) but the increasing
amplitude is caused by positive feedback.
> >For example in a pendulum there is an acceleration
> >(approximately) proportional to the distance from the centre but in
the
> >opposite direction (which I would call negative). If it were in the
> >same direction (positive), the pendulum would not swing but shoot
off
> >like a rocket.
>
> Acceleration from what? Gravity? That's alternately acceleration and
> deceleration and in neither case feedback.
"Deceleration" is not normally used in theoretical work. Acceleration
is a vector in other words it has a magnitude and a direction (not just
a plus / minus sign). What is commonly called deceleration is actually
acceleration which opposes your current velocity. Drive (fly etc) fast
in a tight circle at a constant speed and you will experience what is
commonly called g-force. This is in fact acceleration even though it
does not change your speed. This is because the acceleration is at 90
degrees to your velocity and changes its direction rather than
magnitude.
Back to the pendulum. Yes, the acceleration is caused by gravity. As
the pendulum moves away from its centre it rises a bit. Gravity pulls
it back down which in turn causes it to swing back to the centre. The
further it is from the centre, the greater this force. That is why I
call it feedback, the acceleration is (approximately) proportional to
the distance from the centre. But it is in the opposite direction
which is why I call it negative.
> > Oscillations in electrical circuits are similar, a
> >tuned circuit usually involves an inductor (coil) and a capacitor.
As
> >the charge builds up in the capacitor, it works against the current
> >flow, and like the pendulum, eventually stops and reverses the flow.
>
> And left alone, dampens down in meek obedience to the 2nd law of
> thermodynamics. There's no feedback there either.
It is the same feedback. Current flows into the capacitor and causes
it voltage to increase. The voltage opposes the current and reduces,
and eventually reverses it. Again, the voltage is in the capacitor is
proportional to the amount of current that has flowed in (feedback) and
again it opposes the flow (negative).
That's an ideal circuit with no resistance. A real one would have some
resistance and, without amplification, the oscillation would die out.
When this added to the equation the solution is a sine wave multiplied
by exponential decay.
The maths of the pendulum and the tuned circuit are very similar
(resistance in the circuit is friction in the pendulum). The same
differential equation pops up in both cases.
> Now add an amplifier --transistor or electron tube-- add positive
> feedback, place the coil and capacitor in the circuit. Now you have
an
> tuned oscillator.
I could draw you the circuit diagram if this medium supported it and I
could also give you all the maths. But the significant bit is the
formula that I already gave and you snipped. The negative sign in the
formula is directly related to the feedback being negative.
> >> Where's the feedback in a nuclear explosion?
> >
> >One atom splits and the shards cause a few more to split. The
shards
> >from these few more split a few more each and the reaction grows
> >exponentially (*) until the plutonium is all used up or blown all
over
> >the place.
>
> That's not feedback.
We must have quite different notions of feedback. I use it when the
output of a system is linked to its input. The output of a nuclear
explosion is fragments of plutonium atoms (moving fast and hence
carrying lots of energy). The input is something which prompts the
atoms to split. The fragments from the split atoms (output) cause more
atoms to split (input). This time the feedback is encouraging rather
than discouraging the reaction and hence it is positive. This time,
the snipped formula did not have a negative sign.
> >We are way off topic again.
>
> Yeah, but ain't it fun?
Maybe, provided that it does not bore the rest of the group too much.
--
Seán O'Leathlóbhair
John O'Flaherty
Fast neutrons causing fission, creating more fast neutrons, more
fission. In a nuclear power plant, they have to moderate that positive
feedback with the right amount of negative feedback in the form of
controllable neutron absorbers to keep it from melting down.
--
john
R H Draney
>
>And what's an eme?
The fundamental indivisible particle of vomit...what's a quantum?...r
John O'Flaherty
There, you're calling the tendency of the capacitor and coil to come to
rest negative feedback, which is ok. But what keeps it from dying out is
positive feedback, from the output of the amplifier to the input.
There's no reason why a circuit can't simultaneously employ both
positive and negative feedback.
--
john
John Ings
<quia...@yahoo.com> wrote:
>> Where's the feedback in a nuclear explosion?
>
>Fast neutrons causing fission, creating more fast neutrons, more
>fission.
Yes. But that's not feedBACK. New atoms are involved in each fission.
>In a nuclear power plant, they have to moderate that positive
>feedback with the right amount of negative feedback in the form of
>controllable neutron absorbers to keep it from melting down.
I think you're misusing the word.
feedback, n
a. the process of returning part of the output of a circuit, system,
or device to the input, either to oppose the input (negative
feedback)or to aid the input (positive feedback).
Seán O'Leathlóbhair
Yes, you can have both positive and negative feedback being used in a
complex circuit. I concentrated on the negative in this case because I
believe that it is the explanation of the oscillation. Negative
feedback may cause decay or oscillation depending on whether or not
there is a lag in the effects of the feedback. A lag may cause the
oscillation. But you are right, a carefully controlled amount of
positive feedback is used to prevent the decay. It is similar to the
microphone / speaker problem except in that case, the amount of
feedback is not carefully controlled.
--
Seán O'Leathlóbhair
Seán O'Leathlóbhair
Well John O'F and I seem to agree and I believe that it does fit the
definition that you just quoted.
The neutrons are part of the output of the reaction and it is they that
cause further atoms of the fuel to split. In a bomb, this feedback is
not restrained and it goes bang. In a reactor, it is damped as John
O'F describes and the reaction can be maintained at the desired level.
Why does it matter that new atoms are involved? It is the neutrons
from earlier splits that cause the later splits. Anyway, I would take
a less concrete view, if the level of the output affects the input then
I would call it feedback. The mechanism of the effect does not seem
relevant. The maths is the same regardless.
--
Seán O'Leathlóbhair
John Ings
<jwla...@yahoo.com> wrote:
>> >However I thought that oscillations were an example of negative
>> >feedback.
>>
>> No. Your example of positive feedback from a microphone too near a
>> loudspeaker is an example. That squeal is oscillation.
>
>That is actually a bit of both. The oscillation is caused by negative
>feedback
No, positive feedback.
>(as I believe oscillations usually are)
Wrong.
>> >For example in a pendulum there is an acceleration
>> >(approximately) proportional to the distance from the centre but in
>> >the
>> >opposite direction (which I would call negative). If it were in the
>> >same direction (positive), the pendulum would not swing but shoot
>> >off like a rocket.
>>
>> Acceleration from what? Gravity? That's alternately acceleration and
>> deceleration and in neither case feedback.
>
>"Deceleration" is not normally used in theoretical work. Acceleration
>is a vector in other words it has a magnitude and a direction (not just
>a plus / minus sign). What is commonly called deceleration is actually
>acceleration which opposes your current velocity.
Yes, that's true.
>Back to the pendulum. Yes, the acceleration is caused by gravity. As
>the pendulum moves away from its centre it rises a bit. Gravity pulls
>it back down which in turn causes it to swing back to the centre. The
>further it is from the centre, the greater this force. That is why I
>call it feedback,
Incorrectly. Feedback is when you return output back to the input of
an amplifying device.
>the acceleration is (approximately) proportional to
>the distance from the centre. But it is in the opposite direction
>which is why I call it negative.
That's misusing the word in my opinion.
Can you provide a cite of such usage?
>> > Oscillations in electrical circuits are similar, a
>> >tuned circuit usually involves an inductor (coil) and a capacitor.
>> >As the charge builds up in the capacitor, it works against the current
>> >flow, and like the pendulum, eventually stops and reverses the flow.
>>
>> And left alone, dampens down in meek obedience to the 2nd law of
>> thermodynamics. There's no feedback there either.
>
>It is the same feedback. Current flows into the capacitor and causes
>it voltage to increase. The voltage opposes the current and reduces,
>and eventually reverses it. Again, the voltage is in the capacitor is
>proportional to the amount of current that has flowed in (feedback) and
>again it opposes the flow (negative).
No. Current through the coil builds up an electromagnetic field. The
collapse of that field charges the capacitor, which when the field has
fully collapsed, causes current to flow back through the coil in the
other direction . No feedback is involved. Energy is traded back and
forth between coil and capacitor, converted from electrostatic to
electromagnetic energy of opposite polarities at each oscillation. If
there is no external energy input with positive feedback the
oscillation dies out.
>That's an ideal circuit with no resistance. A real one would have some
>resistance and, without amplification, the oscillation would die out.
>When this added to the equation the solution is a sine wave multiplied
>by exponential decay.
Which would still die out. But:
feedback, n
a. the process of returning part of the output of a circuit, system,
or device to the input, either to oppose the input (negative
feedback)or to aid the input (positive feedback).
>The maths of the pendulum and the tuned circuit are very similar
>(resistance in the circuit is friction in the pendulum). The same
>differential equation pops up in both cases.
And neither involves feedback.
>> Now add an amplifier --transistor or electron tube-- add positive
>> feedback, place the coil and capacitor in the circuit. Now you have
>> a tuned oscillator.
>
>I could draw you the circuit diagram if this medium supported it
http://my.integritynet.com.au/purdic/oscillators.htm
>and I
>could also give you all the maths. But the significant bit is the
>formula that I already gave and you snipped. The negative sign in the
>formula is directly related to the feedback being negative.
No. http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/posfe.html
>> >> Where's the feedback in a nuclear explosion?
>> >
>> >One atom splits and the shards cause a few more to split. The
>> >shards from these few more split a few more each and the reaction grows
>> >exponentially (*) until the plutonium is all used up or blown all
>> >over the place.
>>
>> That's not feedback.
>
>We must have quite different notions of feedback.
Indeed!
> I use it when the output of a system is linked to its input.
That's correct. Now where in a fission reaction is output linked to
input? Or in a pendulum?
> The output of a nuclear
>explosion is fragments of plutonium atoms (moving fast and hence
>carrying lots of energy).
And hitting new atoms of fissionable material. The original atoms are
gone, nothing is fed back to them.
>The input is something which prompts the atoms to split.
>The fragments from the split atoms (output) cause more
>atoms to split
Yes.
>(input).
No.
>This time the feedback is encouraging rather
>than discouraging the reaction and hence it is positive. This time,
>the snipped formula did not have a negative sign.
There is a cynical explanation of the difference between a technician
and an engineer. The engineer it is said, uses math to help him
understand the circuit. The technician on the other hand uses the
circuit to help him understand the math!
I've been an electronics technician for 50 years.
I can't match your math, but I know my oscillators.
>> >We are way off topic again.
>>
>> Yeah, but ain't it fun?
>
>Maybe, provided that it does not bore the rest of the group too much.
We may forgive those who bore us, but never those we bore.
David Wright Sr.
news:cu3341t9r459oi9to...@4ax.com:
(snip)
>> I use it when the output of a system is linked to its input.
>
> That's correct. Now where in a fission reaction is output linked to
> input? Or in a pendulum?
>
>
I agree. Typical confusion of levels of abstraction. 'Knowledge' of the
output is used to feed into the system, but not the 'output' itself, hence,
not feedback by the technical definition.
--
David Wright Sr.
Have you ever stopped to think, and
forgot to start again?
To e-mail me, remove 't' from dwrightsr
John Ings
<jwla...@yahoo.com> wrote:
>Why does it matter that new atoms are involved?
Because of the BACK in feedback.
If it's new atoms, it's feed-forward if anything.
> It is the neutrons
>from earlier splits that cause the later splits.
Yes. A one-way street.
>Anyway, I would take
>a less concrete view, if the level of the output affects the input then
>I would call it feedback.
Yes, I agree there. Now where does that occur in a reactor?
> The mechanism of the effect does not seem
>relevant. The maths is the same regardless.
I think the maths are leading you astray. Puts me in mind of the
math-oriented electronics instructors who will try to tell you current
flows from plate to cathode in an electron tube. Mathematically such
an assumption simplifies a number of equations, but any tech who
encounters that assertion tends to curl his lip and mutter "Yeah
sure!"
Seán O'Leathlóbhair
> >> >However I thought that oscillations were an example of negative
> >> >feedback.
> >>
> >> No. Your example of positive feedback from a microphone too near a
> >> loudspeaker is an example. That squeal is oscillation.
> >
> >That is actually a bit of both. The oscillation is caused by
negative
> >feedback
>
> No, positive feedback.
>
> >(as I believe oscillations usually are)
>
> Wrong.
Ditto.
The dictionary definition that you quote makes no mention of an
amplifying device. You appear to be concentrating on one particular
use of feedback and excluding other reasonable ones.
Also what is output? I think that you are taking it too literally. To
me, the output of the pendulum is its position. This position controls
what the pendulum is going to do next (input). The output affects the
input = feedback.
> >the acceleration is (approximately) proportional to
> >the distance from the centre. But it is in the opposite direction
> >which is why I call it negative.
>
> That's misusing the word in my opinion.
> Can you provide a cite of such usage?
I don't have time to search right now but, if you were to agree that
this was feedback, I can't see that you would call it anything but
negative since the force is trying to bring the system back to its
original state and not away from it. My main reason for calling it
negative is the negative sign that appears in the differential
equation.
Check the physics of a capacitor. As current flows into it, a voltage
builds up that opposes the current. Without the coil, you would also
have negative feedback but you would have the more boring solution of
the current exponentially decaying.
> >That's an ideal circuit with no resistance. A real one would have
some
> >resistance and, without amplification, the oscillation would die
out.
> >When this added to the equation the solution is a sine wave
multiplied
> >by exponential decay.
>
> Which would still die out.
Yes, that is what I meant by "decay".
> But:
>
> feedback, n
> a. the process of returning part of the output of a circuit, system,
> or device to the input, either to oppose the input (negative
> feedback)or to aid the input (positive feedback).
>
> >The maths of the pendulum and the tuned circuit are very similar
> >(resistance in the circuit is friction in the pendulum). The same
> >differential equation pops up in both cases.
>
> And neither involves feedback.
With your very concrete notion of feedback, no. With my more abstract
one that the level of output affects the input then yes.
<snip>
> >> >> Where's the feedback in a nuclear explosion?
> >> >
> >> >One atom splits and the shards cause a few more to split. The
> >> >shards from these few more split a few more each and the reaction
grows
> >> >exponentially (*) until the plutonium is all used up or blown all
> >> >over the place.
> >>
> >> That's not feedback.
> >
> >We must have quite different notions of feedback.
>
> Indeed!
>
> > I use it when the output of a system is linked to its input.
>
> That's correct. Now where in a fission reaction is output linked to
> input? Or in a pendulum?
You seem to agree that my screeching microphone example involves
feedback. How is the output linked to the input in that case? What
comes out of the speaker and back into the microphone?
You snipped without comment my steam engine governor example. Does
that mean that you were happy with it? If so, in what way is the
output fed back?
You seem to want some solid material output to be returned as input but
what is that in the microphone or steam engine examples?
I just want the level of output to affect the level of input. In the
all the examples that we have discussed this applies.
Can you give some more examples, particularly non-electronic ones, that
you would accept involve feedback? In each case, indicate exactly what
is going back.
Is the use of feedback by eBay reasonable? If so, what is going back?
> > The output of a nuclear
> >explosion is fragments of plutonium atoms (moving fast and hence
> >carrying lots of energy).
>
> And hitting new atoms of fissionable material. The original atoms are
> gone, nothing is fed back to them.
OK nothing is going back to the spent fuel but something is going back
to the system (reactor or bomb). Look at the dictionary definition
that you quoted. It is enough that some of the output is returned to
the system. I would see the system as the entire reactor or bomb and
not just the fuel.
> >The input is something which prompts the atoms to split.
> >The fragments from the split atoms (output) cause more
> >atoms to split
>
> Yes.
>
> >(input).
>
> No.
>
> >This time the feedback is encouraging rather
> >than discouraging the reaction and hence it is positive. This time,
> >the snipped formula did not have a negative sign.
>
> There is a cynical explanation of the difference between a technician
> and an engineer. The engineer it is said, uses math to help him
> understand the circuit. The technician on the other hand uses the
> circuit to help him understand the math!
I don't see how it is cynical. It seems just a simple description of
the phenomenon.
I would not describe myself as an engineer or technician.
> I've been an electronics technician for 50 years.
> I can't match your math, but I know my oscillators.
I don't doubt that you know your oscillators.
I once taught electronic engineers in university but I admit that I was
teaching the mathematics modules of their course.
> >> >We are way off topic again.
> >>
> >> Yeah, but ain't it fun?
> >
> >Maybe, provided that it does not bore the rest of the group too
much.
>
> We may forgive those who bore us, but never those we bore.
--
Seán O'Leathlóbhair
Seán O'Leathlóbhair
John Ings wrote:
> On 23 Mar 2005 08:07:51 -0800, "Seán O'Leathlóbhair"
> <jwla...@yahoo.com> wrote:
>
> >Why does it matter that new atoms are involved?
>
> Because of the BACK in feedback.
> If it's new atoms, it's feed-forward if anything.
Look at John Ings' dictionary definition. Something (neutrons) goes
back into the system (reactor or bomb). The fuel by itself is not the
whole system.
> > It is the neutrons
> >from earlier splits that cause the later splits.
>
> Yes. A one-way street.
One split causes another split causes another split . . . Seems
circular to me.
> >Anyway, I would take
> >a less concrete view, if the level of the output affects the input
then
> >I would call it feedback.
>
> Yes, I agree there. Now where does that occur in a reactor?
The more atoms that split the more neutrons are emitted (output).
These more neutrons cause further atoms to split (input). The amount
of neutrons being emitted determines the number of further atoms that
split. Below the critical mass, this does not cause acceleration.
Above the critical mass it does and if you do not control it by
absorbing some of the neutrons, you have a bomb.
> > The mechanism of the effect does not seem
> >relevant. The maths is the same regardless.
>
> I think the maths are leading you astray. Puts me in mind of the
> math-oriented electronics instructors who will try to tell you
current
> flows from plate to cathode in an electron tube. Mathematically such
> an assumption simplifies a number of equations, but any tech who
> encounters that assertion tends to curl his lip and mutter "Yeah
> sure!"
It is an unfortunate piece of history that + and - are the way around
that they are. You need to be consistent in your terminology but
otherwise it makes no difference to the maths which way you say that
the current goes.
Read my reply to John Ing. Answer some of the questions that I asked
him.
--
Seán O'Leathlóbhair
John O'Flaherty
If you choose to look at the mass of fissionable material and its
self-interaction as a system, then you can call the neutron flux
feedback. If you aren't looking at it as a system, then it's not
feedback. There wouldn't have to be feedback to the individual atoms,
which have transmuted in any case.
--
john
John O'Flaherty
If there's a lag that amounts to 180 degrees, then it's really positive
feedback at that frequency. The circuit could still be stabilized by
negative feedback at dc. In a simple l-c oscillator with an amplifier,
there's net positive feedback at the frequency of interest. The
amplifier would increase its output without limit, except that as the
amplitude grows, distortion occurs, effectively decreasing the amount of
energy fed back at the frequency of oscillation. So, an equilibrium is
reached.
--
john
John Ings
<jwla...@yahoo.com> wrote:
>> >> >However I thought that oscillations were an example of negative
>> >> >feedback.
>> >>
>> >> No. Your example of positive feedback from a microphone too near a
>> >> loudspeaker is an example. That squeal is oscillation.
>> >
>> >That is actually a bit of both. The oscillation is caused by
>> >negative feedback
>>
>> No, positive feedback.
>>
>> >(as I believe oscillations usually are)
>>
>> Wrong.
>
>Ditto.
It's a scientific issue, and therefor not a matter of opinion.
Do some research.
>> >Back to the pendulum. Yes, the acceleration is caused by gravity.As
>> >the pendulum moves away from its centre it rises a bit. Gravity
>> >pulls it back down which in turn causes it to swing back to the centre.
>> >The further it is from the centre, the greater this force. That is why
>> >I call it feedback,
>>
>> Incorrectly. Feedback is when you return output back to the input of
>> an amplifying device.
>
>The dictionary definition that you quote makes no mention of an
>amplifying device.
Without amplification (external energy input) the 2nd Law takes over
and everything stops.
>You appear to be concentrating on one particular
>use of feedback and excluding other reasonable ones.
I'll admit to an electronics bias. You extension of the word to
physics seems tenous to me, but perhaps you can back it up.
>Also what is output? I think that you are taking it too literally. To
>me, the output of the pendulum is its position.
To me a pendulum has no output unless it's attached to a clock
escapement.
> This position controls
>what the pendulum is going to do next (input). The output affects the
>input = feedback.
Now you're misusing 'input' and 'output'! :-)
>> >the acceleration is (approximately) proportional to
>> >the distance from the centre. But it is in the opposite direction
>> >which is why I call it negative.
>>
>> That's misusing the word in my opinion.
>> Can you provide a cite of such usage?
>
>I don't have time to search right now but, if you were to agree that
>this was feedback,
Which I don't.
>> Current through the coil builds up an electromagnetic field. The
>> collapse of that field charges the capacitor, which when the field has
>> fully collapsed, causes current to flow back through the coil in the
>> other direction . No feedback is involved. Energy is traded back and
>> forth between coil and capacitor, converted from electrostatic to
>> electromagnetic energy of opposite polarities at each oscillation. If
>> there is no external energy input with positive feedback the
>> oscillation dies out.
>
>Check the physics of a capacitor. As current flows into it, a voltage
>builds up that opposes the current. Without the coil, you would also
>have negative feedback but you would have the more boring solution of
>the current exponentially decaying.
Without the coil you would have no oscillation, even temporarity. Just
a charged capacitor.
>> >The maths of the pendulum and the tuned circuit are very similar
>> >(resistance in the circuit is friction in the pendulum). The same
>> >differential equation pops up in both cases.
>>
>> And neither involves feedback.
>
>With your very concrete notion of feedback, no. With my more abstract
>one that the level of output affects the input then yes.
So as I challenged, find me a cite that shows your 'more abstract'
notion is championed by anyone but yourself.
>> >We must have quite different notions of feedback.
>>
>> Indeed!
>>
>> > I use it when the output of a system is linked to its input.
>>
>> That's correct. Now where in a fission reaction is output linked to
>> input? Or in a pendulum?
>
>You seem to agree that my screeching microphone example involves
>feedback.
Yes.
>How is the output linked to the input in that case?
Speaker output in the form of sound waves reach a microphone which
converts them to electronic signals which are amplified and used to
drive the loudspeaker which converts them to sound waves which reach
the microphone...
> What
>comes out of the speaker and back into the microphone?
Sound energy.
>You snipped without comment my steam engine governor example. Does
>that mean that you were happy with it?
Yes, that's negative feedback.
>If so, in what way is the output fed back?
Rotation of the governor causes weights at the end of its arms to be
thrown outwards by centrifugal force. This action is used to close a
valve reducing the steam being sent to the engine. Output coupled back
to input in negative feedback mode.
>You seem to want some solid material output to be returned as input
No, energy returned as input.
>but what is that in the microphone or steam engine examples?
See above.
>I just want the level of output to affect the level of input.
Yes. Which is what feedback does,
> In the
>all the examples that we have discussed this applies.
Not in the case of free pendulums and nuclear reactions.
>Can you give some more examples, particularly non-electronic ones, that
>you would accept involve feedback?
The steam engine governor. Similar governors on many industrial
internal combustion engines. Water turbine governors.
> In each case, indicate exactly what is going back.
In all cases, energy.
>Is the use of feedback by eBay reasonable? If so, what is going back?
That's the public relations version of feedback that the original
poster was on about. I that case, information. So:
Good company performance leading to satisfied customer resulting in
favorable feedback (Comments) producing good satisfied customer stats
making for good company performance...
>> > The output of a nuclear
>> >explosion is fragments of plutonium atoms (moving fast and hence
>> >carrying lots of energy).
>>
>> And hitting new atoms of fissionable material. The original atoms are
>> gone, nothing is fed back to them.
>
>OK nothing is going back to the spent fuel but something is going back
>to the system
No. Just more coming out.
> (reactor or bomb). Look at the dictionary definition
>that you quoted. It is enough that some of the output is returned to
>the system.
No return is involved. The released particles must hit other unstable
atoms within the system. If they ever escape the system they never
return.
> I would see the system as the entire reactor or bomb and
>not just the fuel.
But what's going on inside it isn't output.
John Ings
<jwla...@yahoo.com> wrote:
>> >Why does it matter that new atoms are involved?
>>
>> Because of the BACK in feedback.
>> If it's new atoms, it's feed-forward if anything.
>
>Look at John Ings' dictionary definition. Something (neutrons) goes
>back into the system (reactor or bomb).
No. Once outside the system, no neutron ever goes back.
To split another atom and enhance the fission, the neutron has to be
inside the system. It is not output, nor is it input. It was always
inside.
>> > It is the neutrons
>> >from earlier splits that cause the later splits.
>>
>> Yes. A one-way street.
>
>One split causes another split causes another split . . . Seems
>circular to me.
Nope. Linear. Always new atoms. Never one that's already split.
>> >Anyway, I would take
>> >a less concrete view, if the level of the output affects the input
>> >then I would call it feedback.
>>
>> Yes, I agree there. Now where does that occur in a reactor?
>
>The more atoms that split the more neutrons are emitted (output).
No. The only output we obtain from a nuclear reactor is radiation and
heat. The principle product is heat, but some of the neutrons are used
to bombard target material within the reactor containment. None are
fed back.
>These more neutrons cause further atoms to split (input).
Again no. They are already within the reactor, they have not been sent
in from outside. They aren't input.
>> > The mechanism of the effect does not seem
>> >relevant. The maths is the same regardless.
>>
>> I think the maths are leading you astray. Puts me in mind of the
>> math-oriented electronics instructors who will try to tell you
>> current flows from plate to cathode in an electron tube. Mathematically
>> such an assumption simplifies a number of equations, but any tech who
>> encounters that assertion tends to curl his lip and mutter "Yeah
>> sure!"
>
>It is an unfortunate piece of history that + and - are the way around
>that they are. You need to be consistent in your terminology but
>otherwise it makes no difference to the maths which way you say that
>the current goes.
Exactly. Math is like that. Which is why a topologist can't tell the
difference between his doughnut and his coffee cup!
>Read my reply to John Ing.
It's John Ings you're replying to. And why do folks keep turning a
perfectly good English name into a Chinese one by dropping the 's'?
Why our family hasn't had anyone drawn and quartered since the Cato
Street Conspiracy!
Peter Moylan
>And what's an eme?
An upside-down ewe.
Actually, I think he meant to say "meme", but was being careful not
to upset Rey.
Peter Moylan
I was thinking about this some more last night, and realised that the
problem is not one of having different definitions in different
disciplines. When people are being careful about definitions, then
everyone agrees that positive feedback means reinforcement of the
current behaviour of the person/system under consideration. So far,
no problem.
The problem arises when vagueness takes over from precision, which
is what tends to happen in certain pseudo-sciences. (Which I won't
name, for fear of offending someone; but rest assured that I'm not
talking about things like psychology.) On top of this, you have the
good/bad value judgment. OK, now you get lots of people with vague
memories of having been taught that if you want to manipulate
others then positive feedback works a lot better than negative
feedback. Now put one of those people into a position of having to
teach in a different discipline, and what pops out is something
along the lines of "positive feedback is defined to be the sort of
feedback that improves a system's behaviour".
This would not be an issue if people stuck to areas they knew
something about. Unfortunately the currently popular economic theories
- popular with politicians, that is - favour throwing people out of
work as quickly and as often as possible, leading many people into
jobs for which they are unsuited.
Peter Moylan
>
>John Ings wrote:
>> On 23 Mar 2005 01:33:41 -0800, "Seán O'Leathlóbhair"
>> <jwla...@yahoo.com> wrote:
>>
>> > But, as
>> >you say, in other contexts, positive feedback is not a good thing.
>>
>> Without positive feedback to cause oscillation no radio or TV could
>> work.
>
>I am not trying to say that positive feedback is always bad either.
>Sometimes good, sometimes bad.
>
>However I thought that oscillations were an example of negative
There's more than one kind of oscillator. A linear or near-linear
oscillator, like your pendulum example, can certainly be viewed
as an example of negative feedback. The sort of oscillator that JI
is talking about is more easily viewed as a system whose linear
approximation has positive feedback and is unstable, but where a
nonlinear saturation limits the instability so that you get a
limit cycle rather than something blowing up.
I used "can be viewed" a couple of times in the above. In fact
there's an element of subjectivity in deciding where the feedback is,
because to do so you have to decide which part is the "system" and
which part is the "feedback". In that sense, the feedback is an
artefact of the mathematical model, rather than being inherent in
the physical system. With a change of variables you can turn
positive feedback into negative feedback, or vice versa. Why?
Because in the physical system it's all lumped together. There
isn't any chalk line giving a block diagram, and there isn't any
box labelled "This part is the feedback".
That's why you can have some people saying that nuclear fission
involves positive feedback, and others saying that it doesn't.
They are both right, but they happen to be using a different choice
of variables to label the inputs, the outputs, and the interactions.
It took me a while to understand this when I was a student. "What do
you mean, that's a feedback resistor? That's just a resistor with
some current flowing through it. I can't see any connection back to
the input." And, in a sense, I was right. The feedback wasn't
inside the resistor. It didn't become visible until you wrote down
the mathematical model. Feedback is a state of mind, because
different observers can have different - but equally correct - models.
Areff
> Actually, I think he meant to say "meme", but was being careful not
> to upset Rey.
Say, where is that fellow?
--
I repeat: Erk, this can't be!
Richard Maurer
In control and systems theory, whose practitioners are mostly
engineers, sometimes mathematicians, and occasionally from other
places such as economics and biology, the terms have a well-understood
meaning. Stripped of the technical jargon, the approximate meanings
are:
- Negative feedback is where some part of the output of a process
is subtracted from the input. The goal of this is, crudely
speaking, to correct any changes in the output by reacting to
its wandering away from where it should be and pushing it back.
- Positive feedback is where a part of the output is _added_ to
the input. The effect of this is to amplify any change in the
output.
I looked in "The Making of America" and "The American Memory"
collections and did not get a hit for "feedback" until a 1992
article on quilts using it in the "comment on" sense.
(There are large gaps in those collections.)
There were a few old Scientific American articles that used
"feed back", but I think this was in a delivery sense (patent
related material is still difficult to comprehend).
I guess that "feedback" started in systems theory,
when they started drawing diagrams.
(And when did systems theory become a separate discipline?)
-- ---------------------------------------------
Richard Maurer To reply, remove half
Sunnyvale, California of a homonym of a synonym for also.
----------------------------------------------------------------------
(word of the day -- manege)
R J Valentine
} "R J Valentine" <r...@smart.net> wrote...
}>
}> [...] Many, if not most, computer people get "simplex"
}> and "half-duplex" switched (as far as I can tell because an ANSI or
}> ISO glossary was based almost to the word on an IBM glossary (which
}> Truly [say, ...] had nothing to do with) that had it backwards. Now
}> there's no convincing people, because they think they understand it
}> (though emes like "simple" and "half" have to be ignored to
}> "understand it").
}
} I'd be surprised if this confusion is as extensive as you say. Let's
} take me as an example -- I think I know what the terms mean, and I'm not
} going to check before exposing my potential confusion. Am I right in
} thinking that both terms label one-way communication channels,
} "half-duplex" being used for those whose direction can be reversed and
} "simplex" for those which can't?
That's the way the computer glossaries have it.
Simplex came first and wasn't called anything until duplex came along
using just about the same equipment, but allowing two-way communications
at the same time by using ground (to get double duty out of a twisted
pair) or another radio frequency to double the capacity. Half-duplex is
simply half a (now called "full") duplex. Standard telecommunications
usage before computer people came along and got it backwards.
ObSparky: Here's another reason not to trust dictionaries.
--
R. J. Valentine <mailto:r...@smart.net>
Steve Hayes
wrote:
>In control and systems theory, whose practitioners are mostly
>engineers, sometimes mathematicians, and occasionally from other
>places such as economics and biology, the terms have a well-understood
>meaning. Stripped of the technical jargon, the approximate meanings
>are:
> - Negative feedback is where some part of the output of a process
> is subtracted from the input. The goal of this is, crudely
> speaking, to correct any changes in the output by reacting to
> its wandering away from where it should be and pushing it back.
> - Positive feedback is where a part of the output is _added_ to
> the input. The effect of this is to amplify any change in the
> output.
>
>you want is negative feedback. Negative feedback stabilises a system
>and gives it various other desirable properties. Positive feedback
>amplifies any imperfections in a system, and can destabilise it.
>In other words, negative feedback is a Good Thing, and positive
>feedback is a Bad Thing.
>
>In the social sciences, however, positive feedback is seen as a
>Good Thing because it helps people's self-esteem; and negative
>feedback is a euphemism for criticism.
>
>come to computer science via a business background, and therefore
>probably had more of a touchy-feely approach to technology than the
>typical computer scientist has. Ergo, confusion over terminology.
And I thought that all that stuff was called "cybernetics".
Do engineers use that term to mean that, or do they use it in some other
arcane sense, or don't they use it at all?
--
Steve Hayes from Tshwane, South Africa
http://www.geocities.com/Athens/7734/stevesig.htm
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk
Charles Riggs
Moylan) wrote:
>Charles Riggs turpitued:
>>When would positive feedback in a social setting be a Bad Thing? That
>>was Peter's point: it is generally a bad thing in electronics, but is
>>usually seen as being helpful when someone gives someone else positive
>>feedback.
...
> On top of this, you have the
>good/bad value judgment. OK, now you get lots of people with vague
>memories of having been taught that if you want to manipulate
>others then positive feedback works a lot better than negative
>feedback. Now put one of those people into a position of having to
>teach in a different discipline, and what pops out is something
>along the lines of "positive feedback is defined to be the sort of
>feedback that improves a system's behaviour".
Good point. I was forgetting the evil controlling people by positive
feedback can bring. Who is to say, in many cases, what the best way to
behave is? People get it wrong many times. What if Mozart's father
thought only sissies played the piano and that skill in kicking a
football should be praised instead?
Charles Riggs
wrote:
>On 23 Mar 2005 01:33:41 -0800, "Seán O'Leathlóbhair"
><jwla...@yahoo.com> wrote:
>
>> But, as
>>you say, in other contexts, positive feedback is not a good thing.
>
>Without positive feedback to cause oscillation no radio or TV could
>work.
A simple crystal radio works without it. Secondly, do all oscillators
require feedback to work?
>>Examples are the screech when a microphone is too near to a speaker and
>>a nuclear explosion.
>
>Where's the feedback in a nuclear explosion?
Nowhere, I suppose, but I can see how the multiplying effect has some
similarities.
Charles Riggs
<ma...@official-totally-reversed.com> wrote:
>"R J Valentine" <r...@smart.net> wrote...
>>
>> [...] Many, if not most, computer people get "simplex"
>> and "half-duplex" switched (as far as I can tell because an ANSI or
>> ISO glossary was based almost to the word on an IBM glossary (which
>> Truly [say, ...] had nothing to do with) that had it backwards. Now
>> there's no convincing people, because they think they understand it
>> (though emes like "simple" and "half" have to be ignored to
>> "understand it").
>
>I'd be surprised if this confusion is as extensive as you say. Let's
>take me as an example -- I think I know what the terms mean, and I'm not
>going to check before exposing my potential confusion. Am I right in
>thinking that both terms label one-way communication channels,
>"half-duplex" being used for those whose direction can be reversed and
>"simplex" for those which can't?
Full-duplex allows both parties to talk at the same time. It applies
to data, as well. Half-duplex, as I understand it, means you talk,
then I talk, then you talk, blah blah blah. Isn't simplex one-way
communication? Coop might know.
>And what's an eme?
Eme, meme, mime, mo, I don't know.
Charles Riggs
Couperus) wrote:
>On Wed, 23 Mar 2005 05:58:35 +0000, Charles Riggs <chriggs@éircom.net>
>wrote:
>
>>
>>When would positive feedback in a social setting be a Bad Thing? That
>>was Peter's point: it is generally a bad thing in electronics, but is
>>usually seen as being helpful when someone gives someone else positive
>>feedback.
>
>Positive feedback is not always a good thing.
>
>If you wantr to encourage a certain behaviour, then yes.
As I said.
>But if you want to discourage a certain behaviour, then
>negative feedback is called for.
Obviously, but that says nothing about positive feedback.
>I like the way you are now paying more attention - keep it up!
Would it disappoint you to learn I don't care what you like?
>I am disappointed that you still blather on at great length
>on matters you know not of. This really does not improve
>your net persona.
Name those matters, then feed us your wisdom concerning them. No-one
is stopping you from doing so, yet we hear nothing.
Linz
"Charles Riggs" <chriggs@éircom.net> wrote in message
news:3vv14193uft6qfe42...@4ax.com...
> When would positive feedback in a social setting be a Bad Thing?
> That was Peter's point: it is generally a bad thing in electronics,
> but is usually seen as being helpful when someone gives someone
> else positive feedback.
When positive feedback in a social setting is a Bad Thing:
Small child is having a tantrum in the supermarket because he wants
sweeties. Mum says no, child yells louder, mum says no, child kicks,
mum gives in. Child has received positive feedback for tantrum.
Ideally child has tantrum, mum says no, child yells louder, mum either
walks away (no audience, negative feedback) or takes child away (no
reward, negative feedback).
I think that's how it works, anyway. I am not a scientist, nor a
social scientist. I am pretty good at carrying a small child under my
arm, though...
Seán O'Leathlóbhair
> On 23 Mar 2005 10:02:03 -0800, "Seán O'Leathlóbhair"
> <jwla...@yahoo.com> wrote:
>
> >> >Why does it matter that new atoms are involved?
> >>
> >> Because of the BACK in feedback.
> >> If it's new atoms, it's feed-forward if anything.
> >
> >Look at John Ings' dictionary definition. Something (neutrons) goes
> >back into the system (reactor or bomb).
>
> No. Once outside the system, no neutron ever goes back.
> To split another atom and enhance the fission, the neutron has to be
> inside the system. It is not output, nor is it input. It was always
> inside.
You seem very fussy about the boundaries of the system. You will not
consider the neutrons as feedback since they don't leave the system.
But look at all the other examples that we discussed. The only one in
which the feedback leaves the system entirely is the microphone case
and that is also the only example of undesired feedback.
Consider a small battery radio. There are some oscillators inside and
although we seem to disagree on the sign, we appear to agree that these
utilise feedback. But nothing leaves the radio and reenters so how is
this feedback by your standards?
Howabout the steam engine? The governor is part of the whole apartus
(system) and hence again nothing has left and reentered the whole
system.
To me the boundary of the system is not so important, just whether
there are loops.
> >> > It is the neutrons
> >> >from earlier splits that cause the later splits.
> >>
> >> Yes. A one-way street.
> >
> >One split causes another split causes another split . . . Seems
> >circular to me.
>
> Nope. Linear. Always new atoms. Never one that's already split.
>
> >> >Anyway, I would take
> >> >a less concrete view, if the level of the output affects the
input
> >> >then I would call it feedback.
> >>
> >> Yes, I agree there. Now where does that occur in a reactor?
> >
> >The more atoms that split the more neutrons are emitted (output).
>
> No. The only output we obtain from a nuclear reactor is radiation and
> heat. The principle product is heat, but some of the neutrons are
used
> to bombard target material within the reactor containment. None are
> fed back.
Neutrons are radiation. The desired product is heat but, desired or
not, neutrons (radiation) are another output.
> >These more neutrons cause further atoms to split (input).
>
> Again no. They are already within the reactor, they have not been
sent
> in from outside. They aren't input.
See comments on boundaries above.
> >> > The mechanism of the effect does not seem
> >> >relevant. The maths is the same regardless.
> >>
> >> I think the maths are leading you astray. Puts me in mind of the
> >> math-oriented electronics instructors who will try to tell you
> >> current flows from plate to cathode in an electron tube.
Mathematically
> >> such an assumption simplifies a number of equations, but any tech
who
> >> encounters that assertion tends to curl his lip and mutter "Yeah
> >> sure!"
> >
> >It is an unfortunate piece of history that + and - are the way
around
> >that they are. You need to be consistent in your terminology but
> >otherwise it makes no difference to the maths which way you say that
> >the current goes.
>
> Exactly. Math is like that. Which is why a topologist can't tell the
> difference between his doughnut and his coffee cup!
Exactly? You say that reversing the current makes the maths simpler
and I say that it makes no difference.
Topology was once my speciality but I never mistook a doughnut for a
coffee cup. Maybe this is because I prefer jam (US jelly) doughnuts.
> >Read my reply to John Ing.
>
> It's John Ings you're replying to. And why do folks keep turning a
> perfectly good English name into a Chinese one by dropping the 's'?
> Why our family hasn't had anyone drawn and quartered since the Cato
> Street Conspiracy!
Double apology here. Both errors were due to excess haste. I mistook
who I was replying to and accidentally let your name go through the
word processor.
The main reason that I classify all of my examples as feedback is that
the maths is similar. The physics may vary, the quantities represented
by the formulae vary, the manner of linkage varies, but the maths is
the same. They are all different physical realisations of the same
maths.
I will try to find time for some research over the weekend but here is
the result of a very quick search. No more posts from me today on this
subject, I am too busy.
http://pespmc1.vub.ac.be/FEEDBACK.html
Note a couple of things:
This article only requires information to be fedback. It does not need
to be anything physical.
Look at the descriptions of positive and negative feedback and their
effects. In particular look at the graphs. Which one oscillates?
--
Seán O'Leathlóbhair
John Ings
wrote:
>>> But, as
>>>you say, in other contexts, positive feedback is not a good thing.
>>
>>Without positive feedback to cause oscillation no radio or TV could
>>work.
>
>A simple crystal radio works without it.
True. Know anybody who uses one except as a toy?
>Secondly, do all oscillators require feedback to work?
Um, that gets a little iffy. Klystrons and twits (travelling wave
tubes) and magnetrons raise questions in that respect.
>>>Examples are the screech when a microphone is too near to a speaker and
>>>a nuclear explosion.
>>
>>Where's the feedback in a nuclear explosion?
>
>Nowhere, I suppose, but I can see how the multiplying effect has some
>similarities.
I only see linearity, not circularity. Let's try an analogy. A pebble
rolling down hill accumulates compatriots until an avalache results.
Feedback? i don't see any.
John Ings
<jwla...@yahoo.com> wrote:
>> >> >Why does it matter that new atoms are involved?
>> >>
>> >> Because of the BACK in feedback.
>> >> If it's new atoms, it's feed-forward if anything.
>> >
>> >Look at John Ings' dictionary definition. Something (neutrons) goes
>> >back into the system (reactor or bomb).
>>
>> No. Once outside the system, no neutron ever goes back.
>> To split another atom and enhance the fission, the neutron has to be
>> inside the system. It is not output, nor is it input. It was always
>> inside.
>
>You seem very fussy about the boundaries of the system.
Yes. I suspect that in a mathematical model, that distinction isn't
important. Our viewpoints prejudice our understanding.
> You will not
>consider the neutrons as feedback since they don't leave the system.
>But look at all the other examples that we discussed. The only one in
>which the feedback leaves the system entirely is the microphone case
>and that is also the only example of undesired feedback.
And that all depends on where you draw the system boundaries.
>Consider a small battery radio. There are some oscillators inside
Just one usually.
> and
>although we seem to disagree on the sign, we appear to agree that these
>utilise feedback. But nothing leaves the radio and reenters so how is
>this feedback by your standards?
That's our boundary problem. In the case of the radio's local
oscillator, the sytem is a transistor. The feedback is from the
transistor's output back to its input. Without that feedback it's just
an amplifier. With positive feedback it's an oscillator.
>Howabout the steam engine? The governor is part of the whole apartus
>(system) and hence again nothing has left and reentered the whole
>system.
Boundary problems again. I see your point.
>To me the boundary of the system is not so important, just whether
>there are loops.
That's the mathmaticians instinctive view I suspect.
>> >> I think the maths are leading you astray. Puts me in mind of the
>> >> math-oriented electronics instructors who will try to tell you
>> >> current flows from plate to cathode in an electron tube.
>> >>Mathematically such an assumption simplifies a number of
>> >>equations, but any tech who
>> >> encounters that assertion tends to curl his lip and mutter "Yeah
>> >> sure!"
>> >
>> >It is an unfortunate piece of history that + and - are the way
>> >around that they are. You need to be consistent in your terminology but
>> >otherwise it makes no difference to the maths which way you say that
>> >the current goes.
>>
>> Exactly. Math is like that.
>Exactly? You say that reversing the current makes the maths simpler
>and I say that it makes no difference.
That was the explanation I was given when I complained bitterly to the
instructor that it was bloody obvious current flows from cathode to
plate in an electron tube, not the other way around. You hit a similar
bit of nonsense in transistor theory when electrons are moving in one
direction in a substrate, and the instructor insists you look at the
atoms the electrons have left and think of the current flow as
'holes' moving in the opposite direction!
>The main reason that I classify all of my examples as feedback is that
>the maths is similar. The physics may vary, the quantities represented
>by the formulae vary, the manner of linkage varies, but the maths is
>the same. They are all different physical realisations of the same
>maths.
>http://pespmc1.vub.ac.be/FEEDBACK.html
>
>Note a couple of things:
>
>This article only requires information to be fedback. It does not need
>to be anything physical.
In my world feedback is always energy.
>Look at the descriptions of positive and negative feedback and their
>effects. In particular look at the graphs. Which one oscillates?
The one with positive feedback. Negative feedback produces dampening
and equilibrium. In an electronic oscilator, you do not see:
"indefinite expansion or explosion (a running away toward infinity)
or total blocking of activities (a running away toward zero)."
The amplifier tries for infinity, but runs out of power, starts to
distort, and reaches a stable state at the frequency dictated by the
associated tank circuit.
I take exception to some of the examples given of positive feedback.
"The examples are numerous: chain reaction, population explosion,
industrial expansion, capital invested at compound interest,
inflation, proliferation of cancer cells."
Industrial expansion, capital invested at compound interest,
inflation-- yes.
Chain reaction, population explosion, proliferation of cancer
cells.-- no.
Tony Cooper
wrote:
>I think that's how it works, anyway. I am not a scientist, nor a
>social scientist. I am pretty good at carrying a small child under my
>arm, though...
Are you good at getting small arms into jackets and sweaters, though?
--
Tony Cooper
Orlando FL
Laura F. Spira
At first reading, that sounds as if you're asking Linz if she's some
sort of gun runner...
--
Laura
(emulate St. George for email)
Linz
"Tony Cooper" <tony_co...@earthlink.net> wrote in message
news:vjk541lul3lu3t88d...@4ax.com...
It helps that YoungBloke is generally helpful but even when he's
pitching a fit, yes, I manage.