Wikipedia:Peer review/Logic/archive1

Logic

This peer review is suspended, because rather far reaching changes were proposed. The page User:Renamed user 4/logic serves as a kind of workshop for generating new content for the revised version of the article, additionally attempting to tackle the other issues raised here and not addressed is welcome. Talk:Logic has a task list of current tasks. --- Charles Stewart(talk) 20:11, 25 February 2006 (UTC)

I think this article is in good shape, and is without equal amongst encyclopedia articles on logic, in terms of not avoiding tackling hard questions as to what the topic is about, comprehensiveness of coverage, and getting into living topics (the two main recent enclyclopedias of philosophy, the Stanford Encyclopedia of Philosophy and the Routledge Encyclopedia of Philosophy each have many articles on topics in logic, but none on logic itself).

That said, there are some concerns with the article:

1. Readability: I've made some recent efforts to try to improve readability, but probably overall the article has several places where the going is too hard going. Help hilighting these is appreciated;
2. I'm worried about topics getting buried: information that should be uppermost getting obscured by other topics. Pointers please to this, and suggestions for changes to organisation that might help this.

That said, I hope there isn't too much to be done here. I've done a fair amount of work on this, CSTAR has done as much, if not more, and the now departed Siroxo has done good work as well, and also many other contributors. CSTAR also did a review of this some months back that is discussed on the talk page. --- Charles Stewart 23:41, 18 November 2005 (UTC)

• Comment: there is a live issue that I didn't notice on the talk page by User:Nahaj on the treatment of inconsistency. --- Charles Stewart 23:53, 18 November 2005 (UTC)

• All in all I find the article a bit too difficult to read. I think the article is crying for examples, schemata and pictures. Why not pictures of famous logicians? But the most important would be pictures or schemata symbolizing paradoxons or the structure of an argumentation or famous syllogisms. I am also missing a picture or a schema presenting basic applications in computer logic. All in all I think the article does not focus enough on pedagogical aspects. Here are some more precise comments of mine -- I hope they shall help you:
  1. typing "define:logic" on google didn't produce anything looking like "study of arguments". This definition seems to me not very great; maybe the authors want to say "study of argumentation". However on the web one finds "study of sound reasoning" or "study of reasoning" which sound much better to me. The definition of argument found on WP is in contradistinction with the one in my dictionary. An argument can be for or against something but not the fact of argumenting.
  2. The sentence ", although the exact definition of logic is a matter of controversy amongst philosophers (see below)" should not be here. The layman does not want to know about such controversy. The lead should be appealing. I suggest: "However several definition of logic are possible." which avoid the use of "controversy".
  3. The sentence "However the subject is grounded, the task of the logician is the same: to advance an account of valid and fallacious inference to allow one to distinguish good from bad arguments." is for me utterly ununderstandable. Is it intended to be understood by the layman?
  4. I think the sentence "even more recently, in computer science" should be changed into "Numerous application in computer science" because logic is surely not a subbranch of computer science.
  5. "The scope of logic can therefore be very large, including reasoning about probability and causality." should be rephrased: what means scope: application? what means "therefore"? Because of "application in computer science". "Therefore" is surely incorrect.
  6. "Here we characterise logic, first by introducing the fundamental ideas about form, then outlining in broad terms some of the most influential rival conceptions of the subject, giving a brief overview of its history and then give an account of its relationship to other science, and finally go on to provide an exposition of some essential concepts." sounds a bit too much like a text book (you know the last paragraph of the introduction section).
  7. The word "inference" should be defined (not only by a link).

  Vb14:57, 22 November 2005 (UTC)

• Thanks for your comments. I've taken the liberty of numbering your points, to make replying easier.
  • I agree that the article needs to be changed with respect to points 3, 5 and 7.
  • Wrt. point 1, I looked at quite a few textbooks and monographs when deciding which definition of logic to use, and I can't explain why dictionaries don't agree with the literature I looked at. There's something similar at work with both arguments and reasons: the content of one's argument may be the same as the content of the reasoning by which one arrives at a conclusion. But argument does appear to be seen as a more fundamental concept than reason for the purposes of logic, because arguments are things that are in the public area that one can point to, whilst reasons appear only indirectly in reports, claims and speculation. It may be worth saying more about the relationship of logic to reason.
    • I think this is a major point. I have learned at school (in French but this is confirmed by my Harrap's French-English dictionary that the word "argument" (as a noun) has the same meaning in French and in English) that an argument is "a reason offered in proof for or against a thing" (Webster's 1987). This is not the same as "argumentation" defined (also like in French) by Webster as "arguing, reasoning". According to those definitions an "argument" is a part of a "argumentation" (which also contains hypotheses, theses and logical relations between the different arguments). You have looked at textbooks and monographs. It is very possible that those focus on some aspect of the topic creating definitions of their own and introduce some jargon for mathematicians or philosophers. I think this article is intended to a broader audience than only these. Vb 09:37, 23 November 2005 (UTC)
  • Point 2: I think the layman should be told about this controversy! I'm all for making the lead section more appealing, but I don't agree with this change.
    • Controversy of these kind exists in almost all mathematics article. There are all the time different definitions corresponding to different contexts or schools. The words "controversy" and "(see below)" can disappear without changing the content of the sentence. I think the suggested change into "However several definition of logic are possible." is exactly attaining this aim. Vb 09:37, 23 November 2005 (UTC)
  • Point 4: I agree with the change, but the place of logic in the academy is unclear. Now I think about it, this is something the article could be plainer about: logic isn't a subdiscipline of philosophy anymore, but is perhaps best seen as an interdisciplinary subject. There may be more logicians working in computer science departments than all other areas combined - so there is some basis for the claim you say is surely not so.
    • I personally still think it is a branch of philosophy and that mathematics is a subbranch of philosophy too. In this sense, some parts of logic pertain to philosophy and others to mathematics but all in all this is a subbranch of philosophy. I would accept logic as an interdisciplinary topic straddling the border between philosophy and mathematics but I would oppose to classifying logic within computer science. That would be like saying mathematics is a subbranch of engineering because most mathematics are nowadays performed in applied science departments. Vb 09:37, 23 November 2005 (UTC)
  • Point 6: It does make sense to indicate the development of the article, though. I remember writing this sentence, and IIRC it was in response to a criticism siroxo made about the article.

I agree that the historical sections should have pictures of famous logicians, and the image repository already has several. I don't see that schemata would make the article easier to follow, though they might serve some iconic role in the article. I'll look over the article for places where examples would be appropriate. --- Charles Stewart 16:54, 22 November 2005 (UTC)

An iconic role would help but I think schemata could help in the context of examples. Vb 09:37, 23 November 2005 (UTC)

Thanks for the additional comments, which will clearly result in improved treatment. I'm rather pressed for time, and doubt I will get another chance to work on the changes before Monday. I'll just make one response now: can you give a good example of where an example would help? The issue with examples for formalised logic is that it takes too much space in such a general article to motivate the formalism to justify the example. Term logic may be an exception, and the article already uses natural language examples in places to justify the need for the topics. --- Charles Stewart 16:24, 23 November 2005 (UTC)

• I haven't had the time to read a lot, but it is not easy going (but logic is also a hard subject). I think the reference to controversy at the start should be kept. Besides, I also think that it does make a subject more appealing if there is some controversy: we want to see blood! However, do define "inference"; I have no idea what it means and how it differs from "argument". I don't understand what "technically grounded" means. I am having a lot of problems with the definition of "purely formal content". Perhaps it will become clearer after a night of sleep, but it would be great if you could give an example there. -- Jitse Niesen (talk) 21:46, 23 November 2005 (UTC)
  • OK, will tackle these three points. I guessed the part about purely formal content would be hard to read, because it was hard to write. I hope I can think of a good way of explaining it. --- Charles Stewart 22:52, 23 November 2005 (UTC)

• Comment: A quick suggestion to improve readability is to use shorter sentences and simpler words as often as possible. Opening material especially benefits from easy, familiar words instead of, say, "fallacious". Likewise, a phrase like "However the subject is grounded, …" is likely to be opaque to many readers, and "to advance an account" is overly fancy. It's a hard discipline for someone with a good education, but it really pays off. --KSmrq^T 22:53, 23 November 2005 (UTC)
  • I've just printed out the article, and I'll go over it offline looking for avoidable technical terminology. --- Charles Stewart 23:24, 23 November 2005 (UTC)

    Eliminating technical terms is a good start, but what I advocate goes beyond that. Consider this paragraph:

    Due to its fundamental role in philosophy, the nature of logic has been the object of intense disputation, and it is not possible to give a clear delineation of the bounds of logic in terms acceptable to all rival viewpoints. Nonetheless, the study of logic has, despite this controversy, been very coherent and technically grounded. Here we characterise logic, first by introducing the fundamental ideas about form, then outlining some of the different schools of thought, as well as giving a brief overview of its history, an account of its relationship to other sciences, and finally an exposition of some of logic's essential concepts.

    I would replace "disputation", "delineation", "technically grounded", "exposition", and perhaps "characterise". I would remove "very" from "very coherent". I would streamline the first sentence. And I would break the last sentence into perhaps three sentences. Philosophers tend to have a horribly opaque writing style; Ernest Hemingway is a better role model. --KSmrq^T 02:52, 24 November 2005 (UTC)

I so agree. Why not say logic is the study of when it is correct to connect sentences by the conjunction "therefore". Dbuckner 21:20, 30 November 2005 (UTC)

Well I don't like it

> I think this article ... is without equal amongst encyclopedia articles on logic

True enough. I can't agree it is very good. It starts out with a poor definition of the subject. I have set up an article Definitions of Logic giving divers definitions of 'Logic'. I think the Penguin definition ("the formal systematic study of the principles of valid inference") is by far the best. It's certainly not just the 'study of arguments' (like pedagogy or rhetoric.

It's also very repetitive. The idea of 'logical form' which is central to most accounts of logic is touched upon all over the place, without being made clear or explicit.

A personal dislike is the mention of Indian, Chinese logic &c. If it had no material impact on the develop of the subject of the main article (as Aristotle did), why mention it except in a separate article. To be sure, the Arabic commentators were material, so include them. But then why omit the significant contribution that the Polish logicians made in the early 20C?

But the main fault is the arbitrary division of the subject. Why divide it (Section 2) into syllogistic, predicate logic (i.e. predicate calculus) – which is a historical division, moreover one which fails to mention what the division really consists of, then modal logic, which is not a historical division, then "deduction and reasoning" bloody hell what is that doing. Then mathematical logic and philosophical logic, which is a cultural and methodological division, then logic and computation which is a clear afterthought.

Why not organise it according to the true division of the subject, referencing the relevant historical distinctions as we go along. E.g.

1. What 'logical form' is - I have some nice definitions in my collection of logic textbooks. How traditional logic viewed if (all/some A is/isn't B), how the predicate calculus views it (there is/isn't some x such that it is/isn't the case that Fx &c). How the idea of the variable or schema is central to formalisation ("The introduction of variables into logic is one of Aristotle's greatest inventions" Lukasiewicz).

2. Truth and falsity – law of contradiction, excluded middle &c

3. Semantics. The intension and extension of terms. Medieval semantics (supposition theory), early modern semantics (propositions connect ideas), modern semantics (model theory).

4. Inference. Medieval theories of inference, consequence &c, early modern and 19c psychologistic theories, modern views. Strict versus material implication, ex falso quodlibet &c

5. Other bits and pieces. Recent developments. Philosophy of language. Computation.

Sorry this is very hurried, I wrote this in 10 minutes as a sketch. I'm happy to make a substantial contribution here.

Dbuckner 21:02, 30 November 2005 (UTC)

• I'm glad to see you are involved here, but I'm not sure exactly what you can mean by "true division of the subject": dividing up the topic is a choice made for expository purposes. I'd very much like to see your treatment of logical form: I found much discussion of the importance of the topic, and very little that really gets to grips with what it is. Some points:

1. Predicate vs. term logic isn't intended as a historical division, in the sense that term logic, as you know, never died. Rather they are intended as rival approaches to formalisation. It's true that the topics in section 2.1-2.3 are not parallel to those in section 2.4-2.7; these were in separate sections but it was felt that the article had too many ramifications

2. Deduction and reasoning is the weakest section of the article; what it is intended to do is to document the role logic plays in education to instill correct reasoning. It could do with a better name: college logic is the term most often used by American educators for this pedagical role, but cuts the topic off from its history.

3. Logic and computation is not an afterthought, in fact I have rewritten it twice, it being my area of specialisation. The problem is rather that the study of logic by those interested in applications in computer science has not has much unity since the early days.

4. The history of logic section, as it stands, is a pretty good refutation of the widely made claim that logic was invented by the Greeks, and we got it from them. It's one of the better sections, I think. I don't think it is a problem that Indian and Chinese logic don't receive further treatment (though a proper treatment of Indian logic on WP would be valuable).

5. The treatment of both semantics and inference is very slight, as it stands, and I'm open to the idea that the article should have more (indeed the lack of a good explanation of what inference is has been . I am concerned that one can't say much here without rolling one's sleeves up and getting technical, more technical that is appropriate for such a broad overview article. If you can persuade me that these topics can be treated in a sufficiently light-handed way, then you will have gone much of the way to persuading me to restructure the article.

6. Your discussion of "other bits and pieces" suggests that your rival structuring for the article isn't going to be any cleaner than the existing one.

I've not had much time for WP this past week, nor will I this week, and much of that time has been taken up with the Carl Hewitt case. --- Charles Stewart 16:57, 1 December 2005 (UTC)

---

Charles - Your points are good ones. But a more detailed look at the article confirms my initial thoughts that the article needs substantial revision. The basic division is wrong (like dividing Americans into Democrats and mid-Westerners), and there are flaws and factual errors, some of them material, in almost every sentence.

For example, the opening section contains no less than five different definitions of the term 'logic'!

• the study of arguments • the account of valid and fallacious inference • allowing one to distinguish good from bad arguments • the investigation and classification of the structure of arguments • the study of fallacies and paradoxes

That is a little too much. Why not something like:

The next paragraph should list examples of the topics. Why does the current article gives as instances 'probably correct reasoning' and 'arguments involving causality'? What is the former? In any case, the list should be a bit longer, and any subject mentioned should get a mention in the body of the article. 'Probably correct reasoning' is mentioned nowhere else. I Googled it and only got 15 hits, the first of which was from the article itself. Thus Google thinks 'Probably correct reasoning' is a topic of major academic importance because of the importance that Google (sometimes wrongly) attaches to things in WP.

I will work on a draft. Is it possible to create a temporary draft page that we could discuss carefully? Perhaps I will put something on my user page. Or does that go against the WP philosophy? My professional work involves the drafting of policy documents, which means the idea of teams of people (each of whom has been carefully selected by examination and interview) working on drafts, discussing each points in detail, checking all points have been discussed, the concept of document control been handed around, the appointment of a co-ordinating editor, the use of a template to which all such documents must conform, the concept of a final draft being signed off by committee then 'locked down' into a 'production' version. The idea of different unrelated unselected people having random control over the production document still seems pretty strange! It does work, but sometimes not so well as others.

On the subject of the Chinese bits, I'm far from persuaded it has anything to do with the development of logic. Even if Boole were influenced, his influence was not direct (he tried to put Aristotle into symbolic form, and did not have the insights of later logicians such as Brentano, Frege, Peirce, Venn &c that led to the important developments. Boole's influence on logic is comparable to John Logie Baird's influence on television (i.e. the idea was great, the implementation was quickly discarded

Also, some logicians would say that genuine 'logic' is restricted to the systematic study of inference forms, which necessarily involves the use of variables or schemata to classify arguments. Aristotle appears to have invented this. I have looked at material on Hindu logic, and the idea of the variable does not seem to feature.

Best wishes --- Dean

---

I agree with most of the points you make. I wouldn't put too much effort into writing a whole, rival logic article right now, rather put together a skeleton of the article so that I can get a clearer sense to what you propose. Use a subpage in your userpsace for the draft (eg. User:Renamed user 4/logic) so that people don't put requests in the middle of your draft, and it has its own talk page. Boole was a key influence on the whole C19 algebraic logic tradition, even if most of the people in this tradition rejected exactly his formulation. I'd say his influence on Pierce is pretty strong. --- Charles Stewart 21:16, 5 December 2005 (UTC)

Hence my comparison to John Logie Baird. He had a certain vision (geddit) but people rejected his way of implementing it. Dbuckner 20:09, 7 December 2005 (UTC)

Draft now available

I see there are still goings on with the Cantor article. Anyway, there is a draft of an article on my talk (not user) page. The idea would be to have a better definition then a section on division, and then scrap the existing section on "formal" logic in favour of something that actually defines the concept of "formal". If you read carefully the current section, you see it doesn't do that, and it is repetitive. Then a short section on semantics. I don't see why this can't be done. Dean

Interesting Sentence

"The formally sophisticated treatment of modern logic apparently descends from the Greek tradition, although it is suggested that the pioneers of Boolean logic were likely aware of Indian logic (Ganeri 2001) but comes not wholly through Europe, but instead comes from the transmission of Aristotelian logic and commentary upon it by Islamic philosophers to Medieval logicians. "

And what does this sentence mean. Is it really a sentence. DB

With parentheses, perhaps the subordination structure of the sentence becomes clearer: ((The formally sophisticated treatment of modern logic apparently descends from the Greek tradition (although it is suggested that the pioneers of Boolean logic were likely aware of Indian logic (Ganeri 2001))) (but comes not wholly through Europe (but instead comes from the transmission of Aristotelian logic and commentary upon it by Islamic philosophers to Medieval logicians)))...

...so it is a sentence, but not a well-written one. It needs reworking, but I'd like to delay that until it becomes clearer about how your proposal will affect the article. --- Charles Stewart 21:48, 7 December 2005 (UTC)

Semantics

I have added a section on semantics in the draft. There is more on modern logic than in the previous draft, thanks to the good work of Dr. Stewart.

I propose to add a section on "controversies in logic" once I have worked through a few more books. I also need to add a reference section. Charles, do you have any references for the work you contributed please.

Dbuckner 16:40, 11 December 2005 (UTC)}