From: "Boyter, Leslie" <[log in to unmask]> Date: Wed, 15 Feb 2012 11:25:51 -0800 Stevan, I am not saying anyone should be denied access to the author's version. I have no problem with OA. In fact, as someone with little to no access to expensive scientific journals, I am happy to be able to read any version I can get. I only have a problem with people implying they read the version of record (by citing it) when they really read and quoted the author's version. I do not necessarily have any particular attachment to the version of record, I just believe in being clear/straightforward/honest about what you actually read. I haven't had a chance to read the full response from Sean Andrews yet, but from what I was able to skim so far I agree with most of it. Citing the actual version you read (especially when there are many iterations) makes the most sense to me. If it was revamped after you read/cited it, at least the reader knows which version you read/quoted. Depending on the infrastructure and/or how the versions are posted/published, the prior versions may or may not be accessible anymore, so I may never be able to read what you read, but at least I know which version it was and why it may or may not differ from the version I am able to access. Yes, I may be idealistic, but I think accuracy in citation is an appropriate ideal. And it's not really that difficult to do (as compared to many other ideals). In fact, it's probably just as easy to cite what you read as it is to cite the version of record. So, you have my blessings to go forth with OA (as if my blessing matters anyway)... just cite what you actually read. ~Leslie -----Original Message----- From: Stevan Harnad <[log in to unmask]> Date: Tue, 14 Feb 2012 06:36:11 +0000 Straightforward question: Since the reason we are discussing authors' refereed, accepted final drafts versus publisher's copy-edited versions of record here is not to compare their relative merits but to determine what Open Access mandates should mandate, do those who point out (correctly) the (possible) shortcomings of the author's draft mean to imply that it is better that would-be users who are denied access to the publisher's version because their institutions cannot afford a subscription should be denied access to the author's version as well, because of the (possible) shortcomings of the author's draft? Because it is as simple as that; all the rest has nothing to do with the practical reality of Open Access (OA) but with scholarly ideals. If we are to reach 100% OA in this decade instead of losing another decade dithering, bickering and digressions, then research funders and research institutions need to mandate author self-archiving. The version with the least publisher restrictions on it is the author's final draft. Over 60% of journals, including most of the top journals, endorse immediate OA self-archiviong of the author's final draft, but not the publisher's version of record. (The rest don't endorse any form of immediate OA.) Are we, in turn, going to endorse this mandate (which -- so far adopted by only 200 institutions -- needs all the help it can get) or are we going to continue debating the relative merits of "that" versus "which"? Stevan Harnad On Mon, 13 Feb 2012, LIBLICENSE wrote: > From: "Boyter, Leslie" <[log in to unmask]> > Date: Mon, 13 Feb 2012 10:47:52 -0800 > > My two cents-- > > I think this is a frightening practice: > >> (4) One can quote from the OA version (which is usually identical to the version-of-record, apart from minor copy-editing). > > > Can you imagine the "minor copy-editing" required for an article that > might completely change the meaning of a sentence or paragraph? I can. > I have seen a ridiculous number of papers written by people that leave > out "minor" words. The author does not see it because (s)he reads the > sentences, paragraphs, etc as (s)he means them (not as they are > actually written). > > Little things, like the word "not" being left out of a sentence, > COMPLETELY change the meaning of a sentence. You can assume the reader > will understand, based on context, that "not" should be there. > However, if the paper is plagued with "minor" errors like this, it > could be entirely misunderstood (or, rather, understood as it is > written and not as it is meant). > > If you are going to cite the version of record, you better read the > version of record. If you are only going to read the OA version, cite > the OA version. At the very least, please do the following: > >> (7) One can (and should) add to the canonical citation the URL of the OA version. > > > Don't get me wrong, I realize editors/proofreaders miss things too (I > have seen a number of edited articles, textbooks, etc containing > errors that cause confusion). I just think if you are going to cite > something, you better read it first. You never know what changes might > be made, minor or not, between the version you are reading and the > version of record. They very well may be important changes. > > > ~Leslie R Boyter