@Howard – Obviously memory doesn’t work exactly in the mathematical sense described here. However, there’s still a loss if some attention must be directed towards [Author name, year].

Oh, and unless my point isn’t clear, I think citations make sense in scientific papers. I’m not saying they are needed in books or blog posts 😛

For papers, you don’t need to memorize the “who said it” part. As long as you go back after writing and do the verification part (which you said was easy), why not just put in the cites while you are verifying?

Citations were a huge benefit to me over this past summer, as I went out to learn general electromagnetics and some specific parts related to permittivity. The citations in papers allowed me to easily build a tree of idea development and to very easily trace sources directly related to what I was doing (instead of sifting through possibly tangentially related search engine results).

I also do think there’s some value in forcing authors to consistently demonstrate a working knowledge of the field. Lists of *relevant* citations provide quick feedback to the reader that the author knows what he/she is talking about, without having to go through the rigor of searching the person’s work.

Lots of these things are more relevant for learning a new field — of course when you already know a topic, you won’t need this sort of information. But, long after you’re gone, it’s nice for future researchers to have an easy to parse record of how ideas were developed over time.

It’s probably also field-specific. I’m in applied math, and so I’m often trying to connect together ideas from engineering/physics/biology with the mathematics.

Maybe another way of making your argument is to make an analogy with things in math like the quadratic formula, or solving basic differential equations. At my point, we all take those things for granted and so we don’t have to cite them. I could see the same being the case for, say, string theorists who are so far down their rabbit hole that much of the background is just assumed to be well-known.

The role of citations varies quite a bit depending on your profession. I’m familiar with the problem of circularity that Jacob remarked on from reviews of scientific literature cited by defense experts and used forensically in court.

@dragoncar is probably coming at this from the perspective of a lawyer. In our work, the authoratative interpretation of a legal issue is not our own but that of an appellate judge memorialized in a published case, which must be cited to rely upon its value as binding precedent. Uncited arguments may be persuasive, but trial court judges hate getting reversed; so they insist on citations. The habit is an ingrained aspect of thinking like a lawyer and, in my case, permeates my approach to any non-fiction liturature. I simply don’t give as much credibility to unfootnoted assertions of fact as I do to those that provide me with a source that I can (but won’t necessarily) consult to see if the work underlying that assertion justifies it.

But much of Jacob’s writing on ERE issues is of a more philosophical sort, where the importance is more on the questions being raised than on the definitiveness of the facts being asserted. I don’t mind the lack of citations in such cases, and I don’t hold it against an author who omits them.

@chenda – Facts can be checked against reality. If the academic discipline doesn’t deal with reality, citations are needed, in particular of the academic discipline mainly deals with citations, i.e., writing reviews of each others reviews. Yet, if the academic discipline deals with reality, one could (and actually should) set up one’s own experiment to verify the claim instead of relying on a citation.

I suspect each discipline has its own attitude towards citations. For example, when I (ex-physicist) see citations rather than explicit arguments I can easily get the impression that the author doesn’t know what he’s talking about and just relies on second-hand knowledge. In other words, I put very little credibility into citations and merely use them as a suggesting as to where to read more.

@hickchick – More like the multiple choice test at the end of an employee training course. Which class of fire extinguisher must be used on an electric fire? A, B, C, D? Similarly, which authors must be mentioned in a sentence stating a fact about … these guys, these guys, or these guys?

@tjt – Why not, you nit picker you? Why not let a cell operate on itself, e.g. A(A).neq.A. Also A(B).neq.B(A).

Incidentally, I initially wrote it down as 190 before I thought that imposing symmetry and reflexivity would be an arbitrary restriction 😉

To write it out, suppose that A(B) is the function that checks if A>B. Then A(B) is not the same as B(A). Without loss of generality, you could also have a function where A(A) is not the same as A, e.g. a logical XOR.

@Maus – NoMoPoMo. I think if the definition matters, one is probably reaching in terms of conclusions anyway. So if the argument is one that says that unemployment dropped by 0.1% from 9.1% to 9.0%, then the methodology matters(*). If the argument is that unemployment is historically high at 9.1% then it does not matter. Another method may say it’s 7.8% and that would still be high.

(*) However, I question the value of a statement like than when other employment numbers say 7.8%.

Overall, it depends on how precise you want an argument to be. From my perspective, the greater the precision required, the more the author takes over the job of thinking for the reader. I often make things intentionally vague so the reader is forced to do at least some thinking. The woodworking books of Roy Underhill does the same. He never gives exact measures which forces the builder to “work a little” when building his own project. I find I learn a lot more from authors who do this and so I try to emulate them. Now my question to you is … maybe some intentional vagueness might be helpful in a common law system … it would allow for more judgment on the part of the judges. Law is after all a technology intended to assist in guiding people’s behavior. It is not an end in itself.

@Jacob
In practice, the vagueness is built into the common law system because no prior case is absolutely on point WRT the disputed issue in the vast majority of cases in current litigation. The whole art of lawyering can be boiled down to the operations of analogy and distinction: explaining how Case 1 is different on grounds a, b & c, and thus inapplicable; whislt Case 2 is similar on grounds x, y & z, and thus controlling. Theoretically, the side that does this most persuasively wins. In actuality, the side that understands best how the trial judge thinks wins. Logic is a necessary but not sufficient cause of victory.

@et – Only approximately. Quantity does not equal quality. It was when I learned about the various tricks used to manipulate those numbers as a postdoc that I pretty much decided that I didn’t want a career in academia.

Also see,

http://en.wikipedia.org/wiki/Bogdanov_Affair
http://en.wikipedia.org/wiki/Sch%C3%B6n_scandal
http://en.wikipedia.org/wiki/Sokal_affair

So which is better? Citations or memory?

I completely disagree. Citations should be used where applicable. There is too much garbage out there touted as truth for citations to be considered irrelevant.

This might be the first time I don’t entirely follow your logic Jacob.

I prefer listening to lectures because I am an auditory learner. Taking notes helps retention. If I just read a book, I miss a lot of stuff. Plus, you can’t raise your hand in the middle of a paragraph and ask the book for clarification.

@Chad said:
> There is too much garbage out there
> touted as truth for citations to
> be considered irrelevant.

As Jacob points out, citations aren’t quite the “holy grail” of facts as they’re made to be. I’ve seen several cases in academia/industry, where people sharing similar “cliques” group together, target a famous conference, and form what is known as “special session”/(SIG) and get their views published (two or three such conferences, and then club the results for a similar big name journal…, four journals, then you have a monograph…etc.). Then it is a case of what Jacob said in the comments (Death by citations). So, personally I do agree with Jacob that some of these things persist by virtue of their tradition.

@Jacob,
P.S: So, when are you going to burst the bubble of that other “holy grail” called IP/Patents (aka, the “bulwark” of “knowledge-based” society/”economy”)? 😎 😛

P.P.S: So, where did you read about “homeoletic”? Citations, please. 😉 :-D!

@jacob Thanks for the references. For the record, I found nothing on the web of goals myself. For Maximum Power I found the same references, and for NAV I did not have the idea to look at the capitalization rate.

Had I known more about the subjects, it would have been easier for me to evaluate what I find through a web search. But not knowing a lot requires building up you own mental model. This may be better than using (trusted) citations, but it also takes a lot more time and effort.

On a related subject, I tried to find a good book about finance and econometrics. There are hundreds out there, and I think probably a lot of good ones. However finding one which teaches you finance relevant to ERE is completely different to finding one which does a good job for your university diploma.

In that sense, maybe citations are not needed, but you need footnotes with pointers to further reading. Otherwise your readers just have to believe you, or start an effort to find out themselves. But starting the effort may (is) not worthwhile always. The links themselves also show how your interpretation is on the subjects.

@simplex – Hence the bibliography in the book.