Here's a small dilemma. I have a few ideas that I am putting into the form of a research proposal. It borrows heavily from the insights of David Ricardo's Theory of Comparative Advantage. Now, the notion of comparative advantage is pretty much common knowledge in economics. However, my proposal is for research in Complex Networks.

Should I cite David Ricardo's original book from 1817 in which he published the idea, without having read it? (I doubt I'll be able to follow the argument easily in Enlightenment-period English) Or should I cite a secondary paper by someone more modern that talks about and analyzes comparative advantage? I would prefer the latter route as I would actually be able to read the material that I'm citing, but then I will not be going to the source of the original idea.

4 Answers 4


You need to read the material you cite. That does not mean you need to read an entire book if the relevant portion is in one or a few chapters, but just plopping in a citation without ever cracking the cover is reference padding and is academic misconduct.

You've been given good advice by J. Fabian Meier about also citing newer material, but you can't cite something you haven't read.

  • 1
    I disagree with the implication that you should necessarily read the material you cite. In my field (theoretical CS), it is common to attribute definitions, theorems, etc., to the original paper where they were introduced, just to give the reader a kind of historical pointer. In many cases, authors do so without any familiarity with the work they are citing, and the citation only acts as a kind of unique identifier. I don't see the harm in this -- it is common practice, and people in the field don't expect authors to have read such references, so I don't see how it would constitute misconduct.
    – a3nm
    Sep 9, 2017 at 21:50
  • @a3nm "Pointers" go in a bibliography. You can even write an annotated bibliography, and many of us have had to do so. Citations are to material you have read, understood, and have assimilated.
    – Bob Brown
    Sep 10, 2017 at 2:09
  • OK, so I think there might be a difference in terminology across fields. In a theoretical CS article, every time you refer to a different article, it is called a "citation" and gets an entry in what we call a "bibiography". There is no material difference between pointers and what you call a citation. Hence, there is no formal expectation for the author to have read the material in citations.
    – a3nm
    Sep 10, 2017 at 7:23
  • 1
    It is common in discrete optimization to cite a bunch of references in the introduction - something like "a similar problem was also considered by [X,Y,Z]". It is usually a waste of time to read all of them thoroughly. I agree, though, that you should not cite a source that you have not looked at or of which you do not know the method or result. Sep 10, 2017 at 7:47
  • 2
    @a3nm There might be no expectation of that in theoretical CS, but if you don't read things you cite then there is a good chance you are propagating inaccurate information, misleading readers, and wasting their time. See my answer here.
    – user72102
    Sep 10, 2017 at 8:27

In general, using the original source is a good thing. But if the original source is very old, and much discussion and development has happened since, it is not advisable to only cite this original source but something more recent. This does not mean that you cannot cite Ricardo in your introduction, but if you go into the details, I would cite a recent, respected work that offers a modern view.


In your place, I would read just enough of Ricardo’s book to support whatever point you're making, and cite it. As Bob Brown wrote, you don’t have to read the whole thing to cite a single claim from a single section.

I’m also a little unsure why you consider Enlightenment-period English so unreadable. Here’s a sample of the prose I encountered when I looked up Ricardo’s 1817 book online:

There are some commodities, the value of which is determined by their scarcity alone. No labour can increase the quantity of such goods, and therefore their value cannot be lowered by an increased supply. Some rare statues and pictures, scarce books and coins, wines of a peculiar quality, which can be made only from grapes grown on a particular soil, of which there is a very limited quantity, are all of this description. Their value is wholly independent of the quantity of labour originally necessary to produce them, and varies with the varying wealth and inclinations of those who are desirous to possess them. These commodities, however, form a very small part of the mass of commodities daily exchanged in the market.

This is easily comprehensible to me even as a non-economist, and a far more clearly written than a lot of present-day scientific articles I’ve read.

  • 3
    Thanks. Yes, I was a little hasty in assuming that the English was hard to read. You're point on not having to read the entire book is well taken. That was more precisely what I was trying to convey in my rather poorly framed question.
    – Joebevo
    Sep 9, 2017 at 15:27

(from the question above)

... a research proposal. It borrows heavily from ...

Always read the original work, unless perhaps you'd have to learn a whole new language. Shakespeare's English is easy enough to understand once you've gotten into it, a text half as old shoudn't be a large problem.

  • 3
    It is usual to cite a lot of papers that you have just skimmed through. It is a waste of time to read everything if there is a high probability that you find nothing interesting in the text. You might say: So if I do not want to read it, don't cite it. But the culture of many research areas demands that you cite 10+ papers if you write one. Sep 9, 2017 at 8:29
  • @J.FabianMeier If you are not prepared to read ten or twenty papers to write one, you should simply not write scientific papers. And why on earth would you cite a paper which has nothing interesting (for whatever you write about) in it?
    – Karl
    Sep 9, 2017 at 14:05
  • My last paper built upon 5 other papers which I thoroughly read. But in the introduction, I need to mention all relevant papers that also considered the problem, even if I knew, after skimming through them, that their method and their results where no help to my research. Sep 10, 2017 at 7:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .