Thursday, October 06, 2005

Quality and quantity of search results from Google

The good news is that Google searching is free; the bad news is that anything can appear, whether true or false. Professor Madison argued the plausibility of an event more than 50 years ago: “[It was] Bell [Labs'] failure to recognize the potential of the transistor in the non-telephone based consumer technology market, which led to Bell’s demanding a foolishly low license fee from Morita.” Now, I don't think for a minute that Bardeen, Brattain, or Shockley, or their managers at Bells Labs, failed to recognize the potential of the transistor (or thought the only application was for hearing aids), but that is a another story.

Elsewhere on the
thread about "law as design", Professor Madison wrote:

I was referring merely to a quick read of articles about the history of Bell Labs that came up on a quick Google search.

And “Google results” shouldn’t be taken disparagingly, even though I don’t have time to give cites and sources; I was looking at e-copies of journal articles — from the technical literature, not the legal literature.

Madison's view may be contrasted with that of Tony Rothman who talks about how the Internet has become "history's largest rumor mill, where false claims get multiplied thousands of times and establish their veracity by weight of numbers." [US 1, Sept 14] The "transistor only for hearing aid" is one such example.

Also, I don't know which technical articles Professor Madison was referencing. However, I did think of law reviews. Therein, all that is necessary to pass cite checking is that the statement appear in the literature. The statement does not have to be true. Thus, a legal writer could write "The world is flat," as long as the writer cited to a reference saying "The world is flat." Further, as evidenced by the Lemley/Moore footnote about Clarke's assumption, the statement in the law review does not have to be backed up by anything at all in the cited article. The authors can just make stuff up.

But, another thing caught my eye: the reference to e-copies of journal articles. The technical journals of the APS and the ACS are not available (for free) as e-copies. While there may be some free technical work, it is, right now, not in the majority of technical literature. Also, Professor Madison might consider "social science" or "economic" work part of the technical literature. Therein, one might enter into the SSRN database including drafts of papers. Selecting from what is available for free on Google can lead to a very limited, and potentially inaccurate, worldview.

Moreover, as Dan Hunter suggested in Walled Gardens, 62 Wash & Lee L. Rev. 607 (2005), the situation in the legal literature is even worse than for technical literature.

Hunter began: The most significant recent development in scholarly publishing is the open-access movement, which seeks to provide free online access to scholarly literature. Though this movement is well developed in scientific and medical disciplines, American law reviews are almost completely unaware of the possibilities of open-access publishing models.

A motivating force in Hunter's article: [A law review editor] informed me that as copyright holder of my articles, the
law review had asked the Social Science Research Network (SSRN) to
remove the draft articles that I had posted there. n3 SSRN is an organization that acts as a free online repository for scholarly papers in the social sciences.

Hunter also wrote:
Scholarly publishing is a quite remarkable market indeed, where the suppliers of the basic product, the authors and editors, provide their content and services for free to commercial publishers, who are then able to extract monopoly rents from the same group of individuals who provided the content in the first place. There are many troubling social costs of this peculiar system:
the public pays multiple times for the scholarly product, n33
researchers from the developing world are incapable of accessing and contributing to scholarly knowledge, n34 and student tuition is unfairly inflated in a fruitless effort to keep research programs and libraries afloat. n35

Returning to Google searches, I have noticed that results on the same search done on different days produce different numbers of hits and a different variety of results. What you learn can depend on "when" you did it. And, you can't really be sure of the accuracy of what you get.

**As a note on ACS journals

In the last week or so, I noticed that the first page of a 1988 article in Journal of Physical Chemistry that Prof. Frenklach and I wrote on carbon clusters (e.g., fullerenes) has become available on the internet:

Other articles by me in JPC are not available at all. The idea that technical articles are routinely, freely available on the internet is simply wrong.

**More on Rothman

In "Everything's Relative," Rothman notes that Edison is widely credited with inventing the light bulb. He didn't. He did invent the first practical version.

**More on quality of internet searches

from Law Dawg Blawg

On Oct 21, 2005, I ran a search on google: +"patent reform" +2795

The first hit for anything I wrote was on page 15 (for "Imagine" in IPT) and the first hit for anything in this IPBiz.blogspot was on page 21 (Cleere). I suspect that the search algorithm of Google has nothing to do with the content of the articles.


Post a Comment

<< Home