ethics - When you reference an article, is it always expected that you have actually read it?

Naturally I would say "of course", but my current situation makes me doubt.

In my paper I'm briefly covering various alternative cryptographic constructions. I am (of course) familiar with all of them, but I have not (nor have the time to) read the full papers I'm actually referencing. The reason for this is that they contain lengthy specifications complemented with cryptanalysis.

I'm in doubt whether it's acceptable to reference said papers without having actually read them. Is it acceptable to do so?

Answer

Let me answer your question by means of an example. In my on-going research I am using a notion called "Schlichting completion" which is originated from a paper written in German by a mathematician called Schlichting. I can't read German, but using a dictionary, I've found some clues about this notion in his paper. So I cited this paper, even though I have only read about one page of it. The reason is sometimes we have to cite a paper, because some fundamental notion has originated from that paper and we have to give credit to the person who invented that notion. Other reasons for citing papers are:

1. 
   

   They have a nice review of the subject and/or contain relatively a comprehensive list of references related to the work.

   
   


2. 
   

   They have done some parallel or complementary works.

   
   


3. 
   
   

   They have some results which are used in my papers.

   
   


4. 
   

   They contain reasons which motivate my work. For example, they ask or suggest a problem which is addressed in my work.

   
   


5. 
   

   They give more examples, applications and/or ideas related to my work.

   
   

etc.

So you do not have to read all contents of a paper before citing it. But make sure it is relevant, useful, some how necessary, interesting and/or important paper with respect to the work you are presenting in your paper. As a final remark, it is always nice to point out where in the paper you are citing is related to the discussion, for example specify the theorem number, the page number, etc.