Re: [ontolog-forum] Dennett on the Darwinism of Memes

[Pavithra <pavithra_kenjige@xxxxxxxxx>]

Dr. Hays ( & Dr.Sowa,  Doug, John Bottoms),

Thank you for the detail explanation about reasoning behind choosing  "thing" ( at your discretion)  as the universal type.
It is a choice to handle unresolved issues at implementation time  by using internal consistency checks vs design time resolution to handle semantic overloads for words like string vs string ( at design time, you can have an additional  field that indicates context of the usage of the word,etc)  and other exhaustive  semantic and grammatical issues or differences  for all practical purposes.   Even though it defeats traditional software engineering principle or best practices to resolve issues at design time vs implementation time,  Languages can get complex, since they are used to express complex universe, with four dimensions or n dimensions,  I understand you all made a decision go with "thing".

My curiosity was more towards Ontology for language or linguistics as a domain ( genetic engineering is a domain or sub-domain for biology) and how it has been handled there.   If we start discussing concepts and semantic of words like " meme",  ontologists have to eventually  model it.

For example SUMO has addressed linguistics.   I wonder how they have implemented it.. 
http://www.ontologyportal.org
The Suggested Upper Merged Ontology (SUMO) and its domain ontologies form the largest formal public ontology in existence today. They are being used for research and applications in search, linguistics and reasoning. SUMO is the only formal ontology that has been mapped to all of the WordNet lexicon. SUMO is written in the SUO-KIF language. SUMO is free and owned by the IEEE. The ontologies that extend SUMO are available under GNU General Public License. Adam Pease is the Technical Editor of SUMO.

Thanks.
Pavithra

---

From: John Bottoms <john@xxxxxxxxxxxxxxxxxxxx>
To: ontolog-forum@xxxxxxxxxxxxxxxx
Sent: Friday, April 12, 2013 2:20 PM
Subject: Re: [ontolog-forum] Dennett on the Darwinism of Memes

PatH,
(see comments below)

On 4/12/2013 4:54 AM, Pat Hayes wrote:
> Pavithra, greetings.
>
> On reading through these emails again, I think I may understand what is bothering you.
>
> I think you have a picture of the semantic/ontology game which divides the world into layers. The base layer is the actual external world, the world of objects. Houses, for example, are in this basic layer. Then language comprises the next layer, which describes the first layer, and words live in this second, language, layer. And perhaps meta-linguistic knowledge, such as disussions of grammatical relationships between words, lives in a third layer lying even higher. But in any case, the layering is strict, and is necessary to avoid paradoxes and perhaps infinite loops in programming, so you object to ideas which seem to violate this clean layered division, by for example using categories (like Doug's and my "thing") which bridge across two or more layers.
>
> If I have this is more or less right, then I would suggest that you might want to consider abandoning this layering idea. It does not really correspond with anything real, and it is not necessary. It reflects a very early approach to avoiding the semantic paradoxes, but this approach was abandoned by mathematicians nearly a century ago in favor of more modern approaches to set theory (which are now considered standard, most notably Zermelo-Fraenkel set theory, dating from 1922) which do not force this rigid hierarchy of levels upon the universe. (Indeed, more recent work has shown that it is quite possible to have set theories which violate the 'layers' idea even more sharply, for example by requiring that all sets have infinitely descending chains of subsets, so there is no "bottom" to the universe. I mention this not to recommend such odd theories, but only to illustrate that the idea that a strict hierachy of levels is somehow necessary, or required in order to keep things consistent, is quite wrong.) The IKL logic developed by myself and others a few years ago is strictly first-order, but it contains its own meta-language: it is able to describe its own propositions and reason about them, and even quantify over them, with complete syntactic freedom, without needing any kind of 'layering' discipline to maintain internal consistency. The propositions are fully-fledged genuine elements of the semantic universe, just like everythng else the logic is describing. (See http://www.ihmc.us/users/phayes/ikl/guide/guide.html ) The RDF semantics uses a simpler version of the same basic device to ensure consistency while allowing such things as classes that contain themselves (rdfs:Class is an element of itself, for example.)
JB: The layering approach traces back to a paper by Edsger W. Dijkstra
entitled "THE Machine" which suggests that operating systems should be
constructed in layers:
http://www.cs.virginia.edu/~zaher/classes/CS656/p341-dijkstra.pdf

Further, Dijkstra points out that in a well-behaved system a layer
passes data to an adjacent layer but does not dip into another's layer's
data. Today, this approach is modified a bit, again for operating
systems. Hierarchical structure is still used, but forward and reverse
tunnels are established to steal data away from a layer using a higher
(temporal) priority.

One can argue that semantic and linguistic processing are types of
operating systems, but I think of them more as applications running
under an operating system. Still, a system that uses an ontology may use
a number of different architectural approaches, and layering might be
useful for some. I don't think we should ban them altogether. I imagine
a sophisticated system would use a combination of layering, data passing
and link types to allow a wide variety of data access.

-John Bottoms
> So, layering is not in fact necessary. I would argue that it is also not desirable. The world is *not* layered in this way, in fact. Language is as much part of the real world as architecture is, and words are just as real as houses. (Contrary to your assertion, below: words *are* objects.) Language is itself described in language just like anything else is: the word "word" describes words just as the word "house" describes houses. And we do in fact violate any kind of layering discipline whenever we talk about reference, because to even describe the idea of reference, we have to say what the referring word is, and what that word refers to, and so mention both 'layers' in one sentence. And when we talk about the business of defining reference itself, as I have just been doing in the previous sentence, we are actually using at least three layers, because this discourse itself is about the *relationship* between language and the world that it describes, which would be located in a higher "level". If you try to keep this kind of discussion separated into rigorously distinct layers of objects, descriptions and meta-descriptions, you will inevitably fail, or become hopelessly confused. Not because the ideas are hard to grasp, but because the layer metaphor is simply inadequate to the task of handling this kind of discussion.
>
> So, to sum up: strictly separated layers of description and meta-description are not in fact required, are artificial, and do not correspond to reality. This layering idea is neither correct nor useful. Abandoning it does not lead to paradoxes or to infinite loops: it just requires us to be a little careful when we make definitions, to check they are internally consistent. But this is a good idea, in any case.
>
> Hope this helps.
>
> Best wishes
>
> Pat
>
>
>
> On Apr 11, 2013, at 3:02 PM, Pavithra wrote:
>
>> Doug,
>>
>> I understand that "Thing" is used in an universal type and everything is a thing, until further defined.
>>
>> The word "house" is an English word that represent an object house,    The word itself is not an object, it is a 5 character string ( which you call a thing) , that represents a building with rooms and doors and roof etc ( which is actually the thing.)
>>
>> Every word represents something in real world,  based on the language.  But it is  part of a language.  Just like number 2 represent 2 things ( we don't know what those two things are, unless specifically mentioned), it is quantitative and part of mathematics.
>>
>> There are actual things, or acts or events and there are language specific words that represents them.  I understand that it  is a matter of  semantics, and  it has been defined and formalized that way.  If we have to change it, we have to define a new class -  language, and subclass - word , types as - noun, verb etc.  But all the double usage ( as in string vs string ) of the same words with multiple context usage etc cause semantic problems.  So I understand that we can not  jump the guns make changes unless  such issues have resolutions.
>>
>> Pavithra
>>
>> From: doug foxvog <doug@xxxxxxxxxx>
>> To: Pavithra <pavithra_kenjige@xxxxxxxxx>
>> Cc: "doug@xxxxxxxxxx" <doug@xxxxxxxxxx>; [ontolog-forum] <ontolog-forum@xxxxxxxxxxxxxxxx>; Pat Hayes <phayes@xxxxxxx>
>> Sent: Thursday, April 11, 2013 4:38 PM
>> Subject: Re: [ontolog-forum] Dennett on the Darwinism of Memes
>>
>> On Wed, April 10, 2013 18:04, Pavithra wrote:
>>> Dr. Pat & Doug,
>>> I remember these discussion and decisions about defining  "Words are
>>> things, just as numbers, patterns and other non-physical
>>> things are. Spoken or written word tokens are *physical* things ",  in
>>> these forums in the past.
>>> But in general discussions in this forum or other contexts, it is not very
>>> intuitive.
>> As a native English speaker, i found it intuitive.  But for the purposes of
>> this forum, you can consider "thing" to be jargon with the defined meaning
>> of the class of whatever can be referred to.
>>
>>>  I wish we used the term "word" itself rather then thing.
>> Most things are not words.  Was your writing of the email i am responding
>> to a word?  Was each bite you took in your last meal a word?  Is your
>> house a word?
>>
>> -- doug
>>
>>> Thanks,
>>> Pavithra Kenjige
>>>
>>>
>>>
>>>
>>> ________________________________
>>>  From: doug foxvog <doug@xxxxxxxxxx>
>>> To: Pavithra <pavithra_kenjige@xxxxxxxxx>; [ontolog-forum]
>>> <ontolog-forum@xxxxxxxxxxxxxxxx>
>>> Cc: Pat Hayes <phayes@xxxxxxx>
>>> Sent: Wednesday, April 10, 2013 3:57 PM
>>> Subject: Re: [ontolog-forum] Dennett on the Darwinism of Memes
>>>
>>> On Wed, April 10, 2013 14:07, Pavithra wrote:
>>>> Dr. Hayes
>>>> Based on wikipedia definition of meme,
>>> Wikipedia should never be considered as a reference.  Use the document
>>> that is the source of whatever Wikipedia claims.  For definitions, go to
>>> a
>>> good dictionary.
>>>
>>>>   it can be modeled as a concept.   It is a social concept.
>>>> http://en.wikipedia.org/wiki/Meme
>>> Wikipedia references the Meriam Webster Dictionary's definition,  "an
>>> idea,
>>> behavior or style that spreads from person to person within a culture."
>>> http://www.merriam-webster.com/dictionary/meme
>>>
>>> Of course, the idea is far more than that.  Dawkins proposed a theory
>>> of how they operate and others have modified and expanded the
>>> theory (creating the field of memetics).  A cultural ontology could have
>>> the basic concept at a broad level and have multiple theories of how
>>> the property of memes in  theory knowledge bases (or ontologies).
>>>
>>>> You can call all "words" as "things".
>>> Needless to say, the word, "thing", has scores of definitions.  The
>>> normal formal ontology use of the term is the one that informs the words
>>> "anything", "something", "everything", and "nothing".  If one says that
>>> "nothing" has a certain property P, but you define words as not being
>>> "things", then a word having the property P would not falsify the claim
>>> that nothing has property P.
>>>
>>> In formal ontologies, the concept "thing" (e.g., Cyc's #$Thing) is the
>>> universal set -- anything that the ontology can refer to (including
>>> #$Thing)
>>> is an instance of it.
>>>
>>>>   And all things as words at your discretion.
>>> Not at all.  Flower petals, animal species, and thing itself are not
>>> words,
>>> although they may have various words or strings of words to denote them.
>>>
>>>> Â But defies the English language, & meaning of the
>>>> word "thing" and how it is described in wikipedia.
>>> Wikipedia's disambiguation page gives the first meaning of the word
>>> "thing"  as "Object (philosophy)", which page states that "Charles S.
>>> Peirce
>>> defines the broad notion of an object as anything that we can think or
>>> talk
>>> about.
>>>
>>> We can think or talk about words, so they would be PhilosopicalObjects.
>>> So would memes (or anything else you mention!).
>>>
>>>> Wikipedia has documented meaning of the word "word" and
>>>> "thing" as follows.
>>>>
>>>> http://en.wikipedia.org/wiki/ThingÂ
>>>>   http://en.wikipedia.org/wiki/Word
>>>> Using definitions of those two words, I can not derive that all words
>>>> are
>>>> things in a logical manner.
>>> ?? Huh?  Are there words that one can not talk about?
>>>
>>>> However I can say that nouns are things.
>>> The referents of verbs, adverbs, and adjectives are also things, in that
>>> they can be talked about.  One can also talk about prepositions, but
>>> if they don't have referents, then their referents are not things.
>>>
>>>> But am not the authority on wikepedia or English language.  So it is at
>>>> your discretion, ( In other words, you are the adviser).
>>> I advise that word is a type of thing, which means that any instance of
>>> word is an instance of thing.  Wikipedia is also a thing.
>>> Â
>>>> However,  In traditional modeling, for example relational and Object
>>>> Oriented world such assumptions leads to many to many relationships and
>>>> causes infinite loops in programming.
>>> Programming languages allow for infinite loops.  Just because the
>>> concept
>>> of thing is an instance of thing, your reasoning engine does not have to
>>> follow the turtles all the way down.  One should program to avoid
>>> infinite
>>> loops.
>>>
>>>> Who is "us" ??  Us is Ontolog group and OWL, UML modelers..
>>> Many in the Ontolog group find OWL very restrictive.
>>>
>>>> About Darwinism, As you said, DNA and genetic engineering did not exist
>>> I presume you mean human knowledge of DNA.
>>>
>>>> at the time of definition.   Felidae
>>>> & Canidae  or Cats and Dogs can not breed an offspring and it is
>>>> fatal if they do so, since they belong to different species
>>> I assume that the word "fatal" referred to being fatal to a continued
>>> line of descent.  There are many cases in which animals of different
>>> species of the same genus can breed to produce sterile offspring.
>>>
>>>>   Who knows what happens in the future or happened millions of
>>>> years ago. I speculate about such things.  I have no proof one way or
>>>> the other at hand.
>>> As species are diverging, members of different subspecies are less
>>> likely to produce fertile (or any) offspring.  It is certainly possible
>>> to have 3 subspecies (S1, S2, and S3) diverging from a parent species
>>> S, such that a member of S1 may sometimes produce a fertile offspring
>>> with a member of S2, and a member of S2 may sometimes produce a
>>> fertile offspring with a member of S3, but no member of S1 may produce
>>> a fertile offspring with a member of S3.
>>>
>>> Sure, no one has proof.  But we have good scientific theories that we
>>> can state and model.
>>>
>>>>   (A Korsak looks like cross breed between a cat & a
>>>>   dog, I may call it a cat ).
>>> The English term is corsac fox (or corsac); its scientific name is
>>> Vulpes corsac.  It is a kind of fox (Vulpes), which is a canine.  You
>>> may call it what you want, but it is not a feline.
>>>
>>>>   I will read or re-read the books that you suggested.Â
>>>>
>>>> Pavithra
>>>> ________________________________
>>>>   From: Pat Hayes <phayes@xxxxxxx>
>>>> To: Pavithra <pavithra_kenjige@xxxxxxxxx>
>>>> Cc: [ontolog-forum] <ontolog-forum@xxxxxxxxxxxxxxxx>
>>>> Sent: Wednesday, April 10, 2013 11:35 AM
>>>> Subject: Re: [ontolog-forum] Dennett on the Darwinism of Memes
>>>>
>>>>
>>>> On Apr 10, 2013, at 8:15 AM, Pavithra wrote:
>>>>
>>>>>
>>>>> John Bottoms:
>>>>>  From a modeling perspective:
>>>>> ÂÂ
>>>>> ·         Languages are first level abstraction of real
>>>>> world
>>>>> ·         Languages are _expression_ of the world, allows
>>>>> us to
>>>>> express and communicate past, present, future, real and imaginary,
>>>>> proven and unproven aspects of the world.
>>>> But they are also in the actual world, and can be studied empirically
>>>> like
>>>> any other phenomenon.
>>>>
>>>>> *   Words are parts of a language.
>>>>> *   Nouns are used to express "things" in English
>>> As are verbs and adjectives.
>>>
>>>>> language.  Things - as in entities.   ( Not all words are things.
>>>>> For
>>>>> example, verbs are words, but not things)
>>>> All words are things. Not all words *describe* things, maybe.
>>> OK.  Words such as prepositions don't describe things.
>>>
>>>>> *   Nouns are a subset of words.
>>>>> *   Memes are ideas / concepts, real or imaginary, proven or
>>>>> unproven.Â
>>>>> Question is :  do  we need to model  meme??
>>>> Who is "we" and what is being "modeled"?
>>>>> My opinion : Concepts can be named with a name and modeled.
>>> OWL users have taken the English word "concept" and made it a jargon
>>> word.  I find such use confusing and suggest avoiding such computer
>>> language-specific use when not referring to OWL.  The Compact Oxford
>>> Dictionary's first definition of 'concept' is "an abstract idea" -- which
>>> seems a good definition to me.  The word comes from Latin 'conceptum' --
>>> "something conceived".
>>>
>>>>>   At present we do not use the verbiage "meme" for it.  Probably we
>>>>> can use the name meme in the future.
>>> The word 'meme' is part of a theory that many do not accept.
>>> There is no need to commit to that theory.  I suggest using the
>>> word only when referring to the theory to which it is attached.
>>>
>>>> My advice would be to only use the term if you have a pretty exact idea
>>>> of
>>>> what it is you are talking about, and document that understanding as
>>>> carefully as you possibly can.
>>>
>>>>> About Darwinism,  Americans use the word Darwinism from a scientific
>>>>> evolution point of view vs theological, god made us, changed us (
>>>>> mutation) etc.
>>>> American scientists use the term the same way other scientists do.
>>>>
>>>>> Maxwell,  & Dr. Steven.
>>>>> Thanks for summarizing my gibberish writing.   ( It was not
>>>>> scientific
>>>>> feed back, it was more of a general discussion)
>>>>> There is natural evolution due to mutation and then there is human
>>>>> intervention for change.
>>>> Until recently, the only intervention available was artificial
>>>> *selection*, which follows the natural process but amplifies the
>>>> effects.
>>>>
>>>>>   Grafting & genetic engineering are human intervention.  My point
>>>>> was
>>>>> Darwinism did not include human intervention, or cross breeding among
>>>>> subspecies.ÂÂ
>>>> Darwin certainly considered cross-breeding and also human intervention
>>>> in
>>>> breeding (eg of dogs and farm animals, which he studied at great length:
>>>> I
>>>> recommend reading his "Origin of Species", it is a very readable work.)
>>>> He
>>>> did not, of course, consider genetic engineering, as genetics had not
>>>> even
>>>> been formulated when he was writing. I suspect he would have been
>>>> delighted and fascinated to have known about genetics and DNA, but he
>>>> did
>>>> not have this pleasure.
>>>>
>>>>> You mentioned that his theory includes cross breeding among
>>>>> subspecies??ÂÂ
>>>> If animals can breed and produce fertile offspring, they are (by
>>>> definition) the same species.
>>>>
>>>>> However the following is not totally proven in all cases and is open
>>>>> for
>>>>> speculation and there are ethical issues about genetic engineering. ( I
>>>>> don;t want to go there)
>>>>>     • crossing between different species is genetically
>>>>> fatal  ..
>>>>> Thanks,
>>>>> Pavithra
>>>>>
>>>>>
>>>>> From: John Bottoms <john@xxxxxxxxxxxxxxxxxxxx>
>>>>> To: ontolog-forum@xxxxxxxxxxxxxxxx
>>>>> Sent: Wednesday, April 10, 2013 6:20 AM
>>>>> Subject: Re: [ontolog-forum] Dennett on the Darwinism of Memes
>>>>>
>>>>> Pavithra,
>>>>>
>>>>> I may not have remembered his wording correctly in the use of "word".
>>>>> Also, it is a difficulty of linguistics that "thing" often gets used
>>>>> when a better selection would be "entity". However, the audience
>>>>> understood the intent of the question. Words come and go and likewise
>>>>> memes come and go. They share some characteristics and there is a
>>>>> shoot-from-the-hip impulse to put them in a lexicon or dictionary.
>>>>>
>>>>> Another view might be that memes are types of propositions that need to
>>>>> be evaluated. They could be classified into "indeterminate" until they
>>>>> are evaluated. Dennett does recognize that memes are "good" or "bad",
>>>>> and I suppose we should accept that they can be resurrected. One theory
>>>>> I have is that the term "meme" applies to atomic entities that have
>>>>> particular attributes or properties that  can be generalized or
>>>>> rationalized. If that is true then we should be able to build
>>>>> classifiers for memes. A question for exploration is whether that
>>>>> property can be understood in a way that makes sense or is useful.
>>>>>
>>>>> Your view of giraffe evolution is referred to as Lamarckian inheritance
>>>>> and it survives today only as a weakened theory.
>>>>> (http://en.wikipedia.org/wiki/Lamarckism)
>>>>>
>>>>> -John Bottoms
>>>>>   Concord, MA USA
>>>>> On 4/9/2013 8:39 PM, Pavithra wrote:
>>>>>> Hello,
>>>>>>
>>>>>> Words are not things.  "Words" representation things if they are
>>>>>> nouns.  memes are ideas that spreads from person to person??
>>>>>>
>>>>>> Darwinism and theory of evolution explains how living organisms evolve
>>>>>> over  few generations according to the needs/usage etc. According
>>>>>> to
>>>>>> him Giraffe has long neck, because they keep stretching their neck to
>>>>>> eat branches and eventually it caused a genetic mutation to  aid
>>>>>> survival -- a process known as "natural selection." These beneficial
>>>>>> mutations are passed on to the next generation.
>>>>>>
>>>>>>   Darwin does not take cross pollination ( for lack of better
>>>>>> word)ÂÂ
>>>>>> of plants and animals and between different species that happens in
>>>>>> one
>>>>>> generation and produce offspring of  blended types  into
>>>>>> consideration.   A Lion and Tiger may have a Liger for a
>>>>>> child.   You
>>>>>> can actually cut a branch of one fruit tree and put it  another
>>>>>> fruit
>>>>>> tree branch stub and tie it up and it may bear the fruit of the first
>>>>>> tree kind..  There is all sorts of intervention that happens to
>>>>>> change
>>>>>> the way species of plants and animal world to evolve  intoÂÂ
>>>>>> something
>>>>>> new and different not only by genetic mutation due to thousands of
>>>>>> years
>>>>>> of  usage or need for survival but due to cross pollination. Â
>>>>>> I know
>>>>>> this is a thesis for genetic decoding not fiction.ÂÂ
>>>>>>
>>>>>> I still have to read the book listed below..
>>>>>>
>>>>>> Pavithra  ÂÂ
>>>>>> ÂÂ
>>>>>>
>>>>>>
>>>>>>
>>>>>> From: John Bottoms <john@xxxxxxxxxxxxxxxxxxxx>
>>>>>> To: [ontolog-forum] <ontolog-forum@xxxxxxxxxxxxxxxx>
>>>>>> Sent: Tuesday, April 9, 2013 5:28 PM
>>>>>> Subject: [ontolog-forum] Dennett on the Darwinism of Memes
>>>>>>
>>>>>> Daniel Dennett's next book will be out in a few weeks and I had the
>>>>>> opportunity to hear him talk about how memes obey the tenets of
>>>>>> Darwinism.
>>>>>>
>>>>>> The title of his book is, "Intuition Pumps and Other Tools for
>>>>>> Thinking".
>>>>>> (not available yet,
>>>>>> http://www.amazon.com/Intuition-Pumps-Other-Tools-Thinking/dp/0393082067)
>>>>>>
>>>>>> His argument starts by asking if words are things. Then he argues that
>>>>>> if words are things then we should consider memes as things also. He
>>>>>> goes on to illustrate that memes follow the basic three principles of
>>>>>> Darwinism.
>>>>>>
>>>>>> His arguments are compelling and I wonder where they belong in the
>>>>>> grand
>>>>>> ontologies of entities. Are memes a new construct, or do memes simply
>>>>>> replicate a known construct?
>>>>>>
>>>>>> -John Bottoms
>>>>>>    FirstStar Systems
>>>>>>    Concord, MA USA
>>
>>
>>
> ------------------------------------------------------------
> 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
>
>
>


_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)