djones@megatest.UUCP (Dave Jones) (12/12/89)
Okay, now *everybody* look up "paradigm" in the dictionary. Don't put it off. Do it now.
budd@mist.cs.orst.edu (Tim Budd) (12/13/89)
In article <11294@goofy.megatest.UUCP> djones@megatest.UUCP (Dave Jones) writes: >Okay, now *everybody* look up "paradigm" in the dictionary. Don't >put it off. Do it now. yes - do that. And when you are totally confused, and wonder what inflectional forms of latin sentences has to do with object oriented programming; then go read Thomas Kuhn's book, ``The Structure of Scientific Revolutions'' (University of Chicago Press, 2nd ed. 1970). No, Kuhn doesn't talk about object oriented programming. But Kuhn is largely responsible for the modern use of the term paradigm in scientific (and those that would be scientific) circles. Unfortunately, most people that use the term don't use it in the Kuhnian sense. I'm afraid, like ``methodology'' (originally the study of methods), it is one of those words that once was defined to mean one thing, and has now been totally redefined by our field.
mjl@cs.rit.edu (12/13/89)
In article <11294@goofy.megatest.UUCP> djones@megatest.UUCP (Dave Jones) writes: >Okay, now *everybody* look up "paradigm" in the dictionary. Don't >put it off. Do it now. Paradigm: a pattern, exemplar, example. (Oxford English Dictionary) Actually, I don't have the OED in my office, but I did find the definition at the start of Robert Floyd's Turing Award Address from 1978 entitled "The Paradigms of Programming." An excellent paper, and one I heartily recommend. Mike Lutz mjl@cs.rit.edu Mike Lutz Rochester Institute of Technology, Rochester NY UUCP: {rutgers,cornell}!rochester!rit!mjl INTERNET: mjlics@ultb.isc.rit.edu
render@m.cs.uiuc.edu (12/13/89)
Written 5:19 pm Dec 11, 1989 by djones@megatest.UUCP: >Okay, now *everybody* look up "paradigm" in the dictionary. Don't >put it off. Do it now. I'll bite. Here's what I get: paradigm: 1. model, pattern. (according to Merriam-Webster) paradigm: 1. a pattern, example or model. (Webster's Unabridged) So, what's your point? Do you think everybody using it incorrectly? Since I think that most are using it in the sense of a "model of programming" or "pattern of development", I don't think that the word is being used all that badly. Still, the word is (I believe) better used to express a physical model, pattern, or example, so it might be better to say that a particular object-oriented system is a paradigm of object-oriented programming. I do not use "object-oriented paradigm" much myself, not because I see it as incorrect, but because I don't think that there is a single, general object- oriented way of doing things. This may change (and doubtless there are some who disagree with me already), but I think discussions on proper terminology should focus more on the danger of trying to standardize without a consensus rather than on fine points of grammar. hal. (Me? I like, "The Tao of Object-Orientation," but the Tao that can be spoken is not the true Tao.)
delft@fwi.uva.nl (Andre van Delft) (12/14/89)
In article <14394@orstcs.CS.ORST.EDU> budd@mist.cs.orst.edu (Tim Budd) writes: >Unfortunately, most people that use the term don't use it in the Kuhnian >sense. I'm afraid, like ``methodology'' (originally the study of methods), >it is one of those words that once was defined to mean one thing, and has >now been totally redefined by our field. I partially agree. ``methodology'' has not been redefined by our field; it still means: the study of methods. It is just that many people in our field don't know what they are talking about, when they say ``methodology''. #:^)| As a consequence of the misuse of this word, *real* methodology is hardly studied at all, in computer science. (Read M.A. Jacksons opinion in Software Development, Prentice Hall) Andre van Delft ***** After all, aren't we just humble programmers ?
djones@megatest.UUCP (Dave Jones) (12/15/89)
From article <77500020@m.cs.uiuc.edu>, by render@m.cs.uiuc.edu: > > Written 5:19 pm Dec 11, 1989 by djones@megatest.UUCP: >>Okay, now *everybody* look up "paradigm" in the dictionary. Don't >>put it off. Do it now. > > I'll bite. Here's what I get: > > paradigm: 1. model, pattern. (according to Merriam-Webster) > > paradigm: 1. a pattern, example or model. (Webster's Unabridged) > > So, what's your point? Do you think everybody using it incorrectly? Yes. Because you read the rest of the line in Merrium-Webster's, you know that it continued, 'esp: an outstandingly clear or typical example or archetype.' An "object-oriented paradigm" would be a program which was written so expertly as to be an exemplary model for other programs of that sort, or one which was so typical of object-oriented programs as to serve as a definitive example. But in any case, it is *a program*, or a model of *a program*. It is not a methodology. It is not a moral or ethical system. It is not a World View. [ The poster goes on to say that "paradigm" more properly applies to a concrete model, and that "object-oriented" is ill-defined. I concur. ] I am aware of the fact that usage determines meaning, and hence the original meaning of "paradigm" is lost to the programming community. It's a pet peeve. (Perhaps I should post this to alt.peeves.) I think the reason it bothers me is that I suspect the new usage is only a pretentious way of obscuring the fact that when we talk about object-oriented things we are discussing very ill-defined notions. Pretend the word "paradigm" were illegal. Then how would you say object-oriented paradigm? (I know, Fen, you would say, 'Object-oriented paradigm,' but just play along, okay?) You might say "object-oriented method", or "object-oriented discipline". But then someone might ask, 'What IS that method?', or, 'What IS that discipline?' But when you say, "object-oriented paradigm", ooooh ... the response is more likely to be a knowing nod. I am reminded of "aspect" and "factor". When I hear a sentence with one of those words in it, an alarm goes off. I ask, what is the substance of this sentence? What objects and actions does it describe? Although both "aspect" and "factor" are perfectly good words, the answer to the question is generally, none whatsoever. In general usage, these words are used simply as syntactic glue to bind words in such a way that the connotations of those words convey an intended value-judgement or emotion, but no concrete meaning. Am I wrong in suspecting "paradigm" also? > ... > > (Me? I like, "The Tao of Object-Orientation," but the Tao that can be > spoken is not the true Tao.) Perfect! (But how come *everybody* thinks he's already enlightened?)
render@m.cs.uiuc.edu (12/16/89)
Written 10:41 pm Dec 14, 1989 by djones@megatest.UUCP: >Because you read the rest of the line in Merrium-Webster's, you know >that it continued, 'esp: an outstandingly clear or typical example >or archetype.' Actually, I have a compact version of Merriam-Webster's and the line that you quote was not there. However, as you say: >An "object-oriented paradigm" would be a program which was >written so expertly as to be an exemplary model for other programs of that >sort, or one which was so typical of object-oriented programs >as to serve as a definitive example. But in any case, it is *a program*, >or a model of *a program*. It is not a methodology. It is not a moral >or ethical system. It is not a World View. Hey, world view. How about "Object-oriented Weltanschauung"? You could impress a lot of people with that. (Silly people, for the most part.) All the points you make (which I have deleted for brevity's sake) are valid. Still, like everything else associated with object-orientation, there is a lot of stuff that will be decided by general consensus. Perhaps it's good to raise the point that "paradigm" as used in this case is improper, but it is more important, I think, to hammer on the fact that there is no general object-oriented way of doing things. Yes, an OO way is evolving, but I don't expect (or want) to see the _IEEE/ANSI Standard Guide to Object-Oriented Systems and Terminology_ for a long while. Okay, I admit you have to refer to "doing things in an object-oriented way" by some simpler phrase. "Object-oriented methodology" is fine with me, just because MY dictionary (Merriam-Webster's, which is, of course, better than all YOUR dictionaries) defines "methodology" as "a body of methods and rules followed in a science of discipline". You can argue about whether CS is a science of discipline, but the word seems to fit the our needs. For those who think this usage of the word "methodology" is a recent corruption, the dictionary is copyrighted 1974, so it's not all that recent. Now if you want to argue grammar, let's talk about "object-orientedness"... hal. "An object never serves the same function as its image or its name." -- Rene Magritte
liberte@m.cs.uiuc.edu (12/16/89)
I think there is a way to understand the phrase "object-oriented paradigm" in the context of the "original" meaning of "paradigm". One interpretation obviously has problems: > /* Written 10:41 pm Dec 14, 1989 by djones@megatest.UUCP in m.cs.uiuc.edu:comp.object */ > From article <77500020@m.cs.uiuc.edu>, by render@m.cs.uiuc.edu: > > paradigm: 1. model, pattern. (according to Merriam-Webster) > > ... 'esp: an outstandingly clear or typical example > or archetype.' An "object-oriented paradigm" would be a program which was > written so expertly as to be an exemplary model for other programs of that > sort, or one which was so typical of object-oriented programs > as to serve as a definitive example. But in any case, it is *a program*, > or a model of *a program*. It is not a methodology. It is not a moral > or ethical system. It is not a World View. However, consider object-oriented *methodology* as one example of possible methodologies. If object-oriented methodology is, in fact, a particularly clear example of possible methodologies, then it is a methodology paradigm. It might more properly be called the "object-oriented methodology paradigm". Why was "methodology" dropped? Perhaps because the term "object-oriented" absorbed the concept of "methodology". I dont think it is because "paradigm" absorbed "methodology". The noun "paradigm" might be appended to other adjectives (derived from nouns, e.g., sailboat transportation paradigm) when the exemplary relationship between the noun and the more abstract class that it is a member of (e.g., modes of transportation) is being emphasized. Might one also say "sailboat paradigm" to mean the transportation paradigm and not some particular sailboat? Here, "paradigm" is certainly not absorbing the concept of "transportation". So, I would say that the meaning of "paradigm" is not being changed, but rather the usage. The class that a thing is a paradigm of is not stated, but implied by context. Dan LaLiberte uiucdcs!liberte liberte@cs.uiuc.edu liberte%a.cs.uiuc.edu@uiucvmd.bitnet
jacob@gore.com (Jacob Gore) (12/16/89)
/ comp.object / djones@megatest.UUCP (Dave Jones) / Dec 14, 1989 / > Pretend the word "paradigm" were > illegal. Then how would you say object-oriented paradigm? "Approach." "Object-oriented approach." As in "Entiry-relationship approach." Jacob -- Jacob Gore Jacob@Gore.Com boulder!gore!jacob
jeff@aiai.ed.ac.uk (Jeff Dalton) (01/04/90)
In article <11345@goofy.megatest.UUCP> djones@megatest.UUCP (Dave Jones) writes: >From article <77500020@m.cs.uiuc.edu>, by render@m.cs.uiuc.edu: >> Written 5:19 pm Dec 11, 1989 by djones@megatest.UUCP: >>>Okay, now *everybody* look up "paradigm" in the dictionary. Don't >>>put it off. Do it now. >> I'll bite. Here's what I get: >> paradigm: 1. model, pattern. (according to Merriam-Webster) >> paradigm: 1. a pattern, example or model. (Webster's Unabridged) >> So, what's your point? Do you think everybody using it incorrectly? >Yes. >Because you read the rest of the line in Merrium-Webster's, you know >that it continued, 'esp: an outstandingly clear or typical example >or archetype.' That's because the dictionary is behind the times. The authors of that dictionary evidently weren't aware of (or decided to ignore) how the meaning of "paradigm" changed after people discovered Kuhn's _Structure of Scientific Revolutions_. Actually, Kuhn's notion of "paradigm" is closely related to the idea of a pattern or model (it's a move up a level of abstraction from the example to what it's an example of), and it turns out that Kuhn used several different meanings of "paradigm" (some of which match the dictionary definitions quoted above). So the new notion of paradigm grew out of the older ones, which is not an entirely unreasonable way for a language to change. >I am aware of the fact that usage determines meaning, and hence the >original meaning of "paradigm" is lost to the programming community. It's >a pet peeve. (Perhaps I should post this to alt.peeves.) I think the reason >it bothers me is that I suspect the new usage is only a pretentious way >of obscuring the fact that when we talk about object-oriented things we are >discussing very ill-defined notions. It's not just the programming community that has now pretty much lost access to the older meanings of "paradigm": all kinds of academics (at least) now use (and overuse) "paradigm" in the new way and seem to be unaware of any of the other meanings. A lot of it is pretentious, just as "methodology" is now often a fancy way to say "method". Nonetheless, I think the new meaning can be used in a reasonable way.