On Oct 7, 2009, at 7:00 PM, Joe Collins wrote: (01)
> Pat Hayes wrote:
>>>
>>> Like functions that are non-computable, the "true" value of a
>>> physical quantity
>>> is non-measurable and therefore not knowable.
>> Even if true, this is irrelevant to ontology engineering. Ontology
>> is not epistemics. The ontology does not set out to describe what
>> can be known. It sets out to describe the reality described by the
>> knowledge. Unless you want to argue that because we cannot know a
>> true value, that therefore no true value exists - a coherent point
>> of view, but not one conducive to communication or normal
>> conceptual description - then please don't put scare quotes around
>> 'true'
>
> I take a third way: I am agnostic on 'true' values.
> Why do we need such an hypothesis, anyway? (02)
Because communication is virtually impossible without this hypothesis,
and because it underlies virtually all of science. Equations like
e=mc2 do not refer to measured values with tolerance ranges and
probabilities of error, they refer to mass and energy. That these
quantities can only be measured to a certain degree of accuracy is
true, but irrelevant to what the equations say. DIfferential
equations, in fact, only make sense when the reality they purport to
describe is assumed to have differentiable fields in it, which are
*inherently* unmeasurable with complete precision. Still, people do
use the equations to make predictions. (03)
> We have known, measured values, which are less likely to be
> fictitious, and they have uncertainty.
>
> In a model, any numerical value you bind to the quantity, Q, that
> suggests that that value is the unique, true value of the object in
> physical reality that Q refers to, is a lie. You can only bind a
> received value to Q, and those values are ultimately received
> through some form of measurement process. (04)
This is a philosophical position, but (with respect) you are not
making a very good philosophical argument in its defense. Rather then
engage in this debate, which I don't have the energy for, all I can
say is, your view is the minority, and ontology-writiing simply isn't
workable without basing it on a more realist perspective. (05)
>
> I'm still not sure there's a consensus on requirements for a UoM
> effort.
> Nevertheless, I'm not saying we need to be able to capture all of
> what is knowable. I am just saying that we should be able to capture
> at least what is actually known. I count among actual knowledge our
> past & current sense perceptions, measurement data, and pure
> mathematics. (06)
I count a lot more than that. But as I say, this isn't the place to
get into a philosophical debate about knowledge and reality. (07)
> The grander sense of reality you refer to (grander meaning large,
> comprehensive and consistent) is achieved by synthetically applying
> abstraction (mathematics) to create models that are consistent with
> our actual knowledge.
>
> You suggest that the purpose of UoM is to "describe the reality
> described by the knowledge", but do you not think it as important,
> if not more important, to describe the knowledge itself? (08)
No. I don't even know how to begin to describe knowledge. AFAIK,
nobody has any theory worth a rat's scrotum about the actual nature of
human knowledge. All I know how to do is to write down theories of the
world. But that is irrelevant here: our task is to construct a UoM
ontology, not to determine importance. (09)
> It seems you are saying that the model we create has more importance
> than the knowledge that proves the validity of the model.
>
> I believe it is antithetical to a scientific perspective to accord
> more importance to any single model of reality than to our
> measurement data upon which the models rest. (010)
Fair enough. I disagree (and I suspect most scientists would), but I'm
happy for you to believe that Im not a scientist. We aren't doing
science in this forum, we are building an ontology. (011)
>
>> Electrons do have a true rest mass, even if we can only measure it
>> to a certain precision.
>
> This notion of a true electron rest mass is not supported by the
> evidence. It is unique to the degree that it can be reliably be
> measured and give the same value repeatedly (to within measurement
> error).
>
> Earlier on this reflector the case of the problematic concept of
> "boiling temperature of water" was pointed out. You may have a nice
> model/theory of phase transitions, but since a (unique) boiling
> temperature can be difficult to measure, reality shows itself to be
> messier than the model. We cannot just assert in the face of the
> evidence that there is a true boiling temperature. (012)
Actually, yes we can. And people do, and the business of the world
depends on such assertions. Not every user of units is a scientist
with a non-realist metaphysics. (013)
> We cannot predict when evidence contrary to any expectations we hold
> will arise as we refine our measurement processes, not even for an
> electron's mass. The quantum theory holds that there is an essential
> uncertainty in the concept, not just a measurement uncertainty. (014)
>
>> We still however assume that they exist, and our science and
>> engineering and trade rules and laws and our ordinary communication
>> all are based on this presumption. So our ontologies should speak
>> of these things that we all assume to exist.
>
> I don't see how they are based on the hypothesis of 'true' values. (015)
They are when they ignore questions of how quantities are measured,
accuracy, etc., and simply refer to values, which they often do. The
weight of my car is given in kilos. It does not come with an error
estimate or a tolerance interval: it is expressed simply as a numeral.
Now, this numeral, when used in an ontology, must refer to something.
What is certainly *seems* to refer to is a weight, specified using a
scale calibrated in kilograms. An *actual* weight, *the* weight of my
car. If there is no such thing as the actual weight, then I cannot
refer to it, and must use some other way to specify what '5647 kilos'
means. So, what, in your view, are we to say about such language? That
is is simply wrong, and should be rejected? Or that it is an
abbreviation of a much more complicated assertion not about actual
weights at all, but rather about measurements of weight, perhaps
referring to the apparatus that was used to measure the weight of my
car? (But what if there is no such apparatus, and the weight has been
calculated in some way rather than measured? Is it then no longer a
weight, but only a pure-mathematical entailment?) Or that a tolerance
interval must exist, even if we do not know what it is, so this must
be expressed as an existential claim about an unknown tolerance
interval? (But that could be any interval: surely we want to say more
about it, perhaps that is a 'reasonable' such tolerance interval for
things of this kind - so we need to talk about what kind of thing the
car is, and have an overarching ontology of reasonableness of
measurements for certain purposes...) But all I wanted to say, and all
that the shipping company needed to know, was how much my car weighs,
expressed in kilos. (016)
There is a deeper argument, in any case. Take a tolerance interval as
an example. You want to say that 4.0 'really' refers to the interval
[3.95, 4.05] , because exact quantities may not even exist. But that
interval expression itself uses exact quantities, viz. 3.95 and 4.05.
Indeed, given your rules here, these are more exact than the original,
having two decimal places: so each of these must really be an
interval: 3.95 is [3.945, 3.955]. And that in turn is [[3.9445 3.9455]
[3.9545 3.9555]] And so on ad infinitum. Now, one can indeed make
mathematical sense of these infinite nested-interval structures, but
its a lot more complicated than the authors of the SI system had in
mind, I am willing to bet. The point being that at some point in this
descent, we usually want to just toss in the towel and say, OK, I'm
referring *exactly* here, just using a numeral to refer to an *actual
number*, a *point* on the rational continuum. And if we can do that
anywhere, then we can do it right at the beginning. (017)
> They are based on values measured against defined standards, the
> essence of the SI & VIM.
>
> Some people understand the Fahrenheit and Celsius temperature scales
> to be based upon the melting and boiling points of water. That may
> have been true in the past, but they are currently defined in terms
> of the more precisely and reliably measured triple-point of water
> (referenced to absolute zero). One reason, I presume, is that the
> boiling point failed to be a true, consistently measurable value.
> The choice of standards for units are largely predicated on the
> degree of precision to which we can measure certain quantities. The
> standards are subject to being changed when we can improve our
> overall precision in measuring those quantities by different methods
> and different defined quantities. It is, for example, expected that
> the standard for mass will soon change. (018)
No argument there. But this is because the standards, indeed, do rely
on abilities to measure accurately and repeatedly. But I'm not
interested in formalizing measurement standards themselves. (019)
>
> Trade rules and laws, when sensible, defer to scientific,
> engineering, and mathematical knowledge (and don't, for example,
> legislate the value of pi or the nature of reality).
>
>>>
>>> I think the more important questions are "Must it be done?" and
>>> "Can it be
>>> done?" rather than "How long will it take?".
>>>
>>> My answers are "Yes", "Probably", and "I don't know", respectively
>> and mine are: no, not in the near future, and God alone knows, but
>> people have been trying for several decades and havn't got very far.
>
> If God alone can have an idea of when a computational model of a
> fairly well-defined representation of uncertainty can be achieved,
> how can any mortal ever hope to make progress on the ontology
> engineering task to "describe the reality described by the knowledge".
>
> To me, the latter sounds ill-defined, far more complex, hence
> enormously more difficult. (020)
That may be because you understand it to mean, describe ALL of said
reality. But that is not the intention. By and large, naive-realist
accounts of the world are much simpler than operationalist accounts.
Just take the example of the weight of my car. Most of human thought,
language and business operates on the naive presumption that the
physical world is real and that one can assert facts about it. I know
that the Copenhagen interpretation of QM has caused some scientists to
reject this naive vision, but (a) that is now only one possible QM
interpretation, and I prefer others which take a more robust view of
reality and avoid Bohr's duality mysticism, eg the transactional
interpretation, and (b) I defy anyone to try describing any part of
the mesoscale everyday world without making some robustly classical
naive-realist assertions. Your example of the bread and milk at the
supermarket is a good example, in fact. Surely, if there are no actual
electrons, there are no actual loaves of bread. So what your wife
should have said is, perform actions which result, with a probability
greater than 0.99, in sense perceptions of the kind naively described
as 'having a loaf of bread in your shopping basket', or some such
formulation. Or maybe it should be 0.999, giving you a one-in-a-
thousand chance of mistaking a bottle of ketchup for a loaf of bread. (021)
>
>>>> Yes, in the abstract, 'uncertainty' is well-defined. It is the
>>>> expression of
>>>> uncertainty that is interesting. Presumably we would include the
>>>> concept so
>>>> that uncertainties can be expressed. Would you be happy with a
>>>> 'quantity-value has-uncertainty simple-quantity-value' property
>>>> (with the
>>>> basic SI definition = the maximum amount by which 'quantity-
>>>> value' may differ
>>>> from the actual value)? What about conditions, e.g. "at standard
>>>> temperature
>>>> and pressure"?
>>>
>>> So, when is something completely unambiguous? Almost never.
>> ? Why are you talking about ambiguity? That is yet another distinct
>> topic (apart from probability and tolerance).
>
> If I specify a quantity as "4.0 newton meters", some may say that is
> an ambiguous quantity, since I have not specified the Kind, e.g.,
> torque or energy.
> I say it is also ambiguous if you do not interpret or specify an
> uncertainty, for example, if "4.0" is not interpreted as the
> interval [3.95, 4.05] (SI standard practice interprets the
> interval). This raises (in my mind) the question, how much more do I
> need to specify to be unambiguous? I believe the answer is
> subjective, or based on requirements.
> My answer (currently) in order of importance is
> (nominal numerical value, coherent derived unit value, uncertainty
> value, derived quantity value, . . .)
>
>>> How much qualification is enough? As much as it takes for the user
>>> to be happy.
>>> Does the user always specify what it takes to make him happy? No.
>>> I think that an expression of uncertainty is a first level of
>>> qualification.
>
>
>>> So, my wife hands me a grocery list. It says "milk, bread, eggs".
>> OK, already this example is irrelevant. Your problem here is
>> interpreting natural-language in a social context. This is about as
>> far as it is possible to get from anything to to with ontology
>> engineering.
>
> This is meant to be a hypothetical use case.
> Please ignore any social-context or natural language issues. The
> context is meant to demonstrate that uncertainty is an everyday
> concern, even if it is not explicitly expressed.
> It is also meant to illustrate how resistant users can be in
> providing unambiguous specifications of what they want. In this
> case, even the nominal numerical value is unspecified.
>
> Assume, instead, you are engineering an agent to fulfill a grocery
> order. What does agent need to 'know' to satisfactorily fulfill the
> request? What is an unambiguous specification? (022)
Enough to get the task done. In this case, how many gallons (pints in
the UK) of milk, loaves of bread, etc.. Nothing at all about
probabilities of error or tolerances, however. Still less do we need
to be in any doubt about the ultimate reality of bread and milk, or
the quantum uncertainty in locating them on the supermarket shelf. (023)
Pat (024)
>
>
> R/jbc
>
>>> Suffice it to
>>> say that my wife is not easily trained to my expectations: she
>>> never specifies
>>> amounts, let alone a range of acceptable quantities. I do like to
>>> keep her
>>> happy, though, and, as important, I don't want to have to repeat
>>> the shopping
>>> trip daily.
>>> I have learned that, having 3 kids, a quart of milk is
>>> insufficient, and ten
>>> gallons is excessive. I've even narrowed it down better than that.
>>> Perhaps she
>>> has trained me to her expectations.
>>>
>>> I am happy to report, however, that, even after untold man-years
>>> of effort,
>>> computers are dumber than I am (stay tuned?).
>>> I'm pretty sure that if computers are ever able to deliver on such
>>> a request,
>>> then to keep my wife happy they will have to have a rudimentary
>>> capability to
>>> process uncertainty. Until they can, I still am useful. Perhaps I
>>> am trying to
>>> put myself out of a job?
> --
> _______________________________
> Joseph B. Collins, Ph.D.
> Code 5583, Adv. Info. Tech.
> Naval Research Laboratory
> Washington, DC 20375
> (202) 404-7041
> (202) 767-1122 (fax)
> B34, R221C
> _______________________________
> (025)
------------------------------------------------------------
IHMC (850)434 8903 or (650)494 3973
40 South Alcaniz St. (850)202 4416 office
Pensacola (850)202 4440 fax
FL 32502 (850)291 0667 mobile
phayesAT-SIGNihmc.us http://www.ihmc.us/users/phayes (026)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uom-ontology-std/
Subscribe: mailto:uom-ontology-std-join@xxxxxxxxxxxxxxxx
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/uom-ontology-std/
Shared Files: http://ontolog.cim3.net/file/work/UoM/
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UoM_Ontology_Standard (027)
|