Charles Petzold

Wednesday is Library Day!

October 3, 2007
New York, N.Y.

When I'm writing a programming book, Wednesday is pretty much like any other weekday. I'm coding, or writing about my code, or a little of both.

But when I'm working on a somewhat different kind of book — and that's the case now, although I can't go into details until my publisher gives me permission to blog about it — then Wednesday is very often Library Day. I get out of the house and go to the library, preferably without any electronics accompanying me.

The New York Public Library system (web site at consists of Circulating libraries and Research libraries. There are about 40 Circulating libraries in the borough of Manhattan alone. These are much like the libraries in smaller cities and towns. You can read recent magazines and newspapers, browse the stacks, and take books home to read if you promise to return them.

The NYPL has only four Research libraries. They are:

The Research libraries have many more books and periodicals than the typical library, but with a catch: You can't take the stuff off the premises.

To some people, this sounds horribly restrictive, but there's obviously a good reason for it. Many of the books and periodicals can't be easily replaced if they're lost. Also, if you need to look at something, the only way it can be "checked out" is if somebody got to the library earlier that day than you and wanted the same book! (Sometimes stuff is missing or mis-shelved but that's another issue altogether.)

I actually find the pressure of reading the book or article at the library very helpful. It "focuses the mind," as they say. Sometimes I'll photocopy some pages for taking home, but some books are too fragile to be copied.

To use the Research libraries, the first step is visiting the Research Libraries Online Catalog or CATNYP. You can do this at the library or conveniently at home. Here you can search by title, author, subject, or periodical. It's also possible to do a search by "word" but these are words appearing in the catalog information for the book and not the book itself.

The catalog entry for the book or periodical will tell you in what Research library the book is located. For many of the books and periodicals I use, that's usually SIBL or the Humanities & Social Sciences Library. Although the Humanities & Social Sciences Library has several rooms devoted to particular subjects (as well as a microfilm room) for most books you go to Room 315.

From the 3rd floor hallway, you enter Room 315 through the Catalog Room. Here is where the walls were once blanketed with the cabinets containing the card catalog in long drawers.* The card catalog no longer exists. Instead, the Catalog Room has a bunch of computers that let you access CATNYP. Otherwise, the process is the pretty much the same as it was when I first entered Room 315 as a teenager in the 1960s: For each book you want, you fill out a little "call slip" with the catalog number, the title, the author, and your name and address. Then you turn your call slips over to a person behind the desk. (Only three at a time, please.) That person writes a number on them, and gives you the carbon. The original goes in a metal tube and enters a pneumatic system where it is delivered to the workers in the stacks. The stacks are not open to the public. I once saw a movie where the workers who pulled books from the stacks glided around on roller skates, but I don't know if that's true.

After you turn in your call slips, you go into the Main Reading Room, which is an enormous and gorgeous room with two wings, and lot of tables and chairs (and yes, Ethernet ports).

It used to be that you went to the South Wing if the number written on the call slip was even, and the North Wing for odd numbers, but these days all the numbers are even. You go into the South Wing and wait for your number to appear on the board, and then you can pick up your books (they come by dumbwaiter) and then just sit and read.

A couple years ago, the NYPL began requiring people who use the Research libray to have a special Access card. The card is easy enough to obtain — you can fill out an application online and then get your photo snapped right in Room 315 — and all they do is scan the barcode on the card and the barcode on the book so they have a better record of who last had their hands on what book.

While waiting for your number to come up on the board, you might want to roam the perimeter of the Main Reading Room, which contains shelves that you are allowed to browse with many types of reference books, standard histories, biographies, etc., both new and old. (Want to take a look at the famous Encyclop├ędie created by Denis Diderot and Jean d'Alembert in 18th century France? It's right there on the shelves.)

The Science, Industry & Business Library (SIBL) opened in 1995. To use the Research library at SIBL, you go downstairs. (The ground floor is a Circulating library.) Otherwise, the process is pretty much the same: Fill out a call slip, turn it in, and wait for your number to appear, but it's all in the same room and the same curvy desk.

If you have a question, you can ask one of the professional librarians. They are so nice and so helpful you sometimes feel like buying them dinner. I'm sure they get marriage proposals all the time. It's their job to be helpful, of course, and they're often quite concerned when you can't get your hands on something you need, and they never ask you why you want to look at some particularly obscure book or periodical. (I once tried to explain why I needed to see the first edition of some book rather than the second edition, and the librarian cut me off saying "But of course.") The staff at the NYPL consists of some of the nicest people in the City, and hence, in the world.

Today was a two-library day. I first walked from my apartment to SIBL, and then to the Humanities & Social Sciences Library, and then back home.

The subjects of interest today centered around the 1943 article "A Logical Calculus of the Ideas Immanent in Nervous Activity" by Warren McCulloch and Walter Pitts, published in the Bulletin of Mathematical Biophysics. This article was the first to describe neurons as little logic gates in a network, and effectively was one of the kick-starters of the whole cybernetics movement. I'd already snagged the article itself in an earlier visit. Now I was interested in some related items.

One of the articles I wanted was a profile of Walter Pitts published in a recent issue of the journal Perspectives in Biology and Medicine. This issue was available online through Project Muse, but only from a library IP address. In the same issue, I was thrilled to discover a related article by Michael A. Arbib entitled "Warren McCulloch's Search for the Logic of the Nervous System," so I downloaded PDFs of both articles for convenient perusal at home.

I needed the actual printed issue of a recent Journal of the History of Biology to read Tara H. Abraham's article "Nicholas Rashevsky's Mathematical Biophysics" about the Ukranian emigre who was at the University of Chicago for many years, and who inspired several students and associates to create mathematical models of biological functions, and who founded the Bulletin of Mathematical Biophysics in which the McCulloch & Pitts article was published.

Up to the Humanities & Social Sciences LIbrary, where I got a recent issue of the Journal of the History of the Behavioral Sciences with another article by Tara H. Abraham, this one entitled "(Physio)logical Circuits: The Intellectual Origins of the McCulloch-Pitts Neural Networks." I took notes on most of it, but then broke down and photocopied the wealth of material in the final half-dozen pages, including what appears to be a very interesting bibliography.

I also had the opportunity to look at a book entitled The Hixon Symposium: Cerebral Mechanisms in Behavior. Although the book was published in 1951, the actual symposium was held in 1948, and the lead article is — I believe — John von Neumann's very first publication on the subject of natural and artifical automata. I read some of it, but the book was so interesting and obviously seminal, so I decided I needed to find a copy for purchase from ABE.

Although the Research libraries are large, they are not large enough. Some stuff is stored in a warehouse in New Jersey and it takes a day or two to be delivered to the library. At SIBL I turned in a slip for two issues of the journal History and Philosophy of Logic. I'll be able to look at those next Wednesday.

How much of this research will actually get into the book I'm working on now? I'm not quite sure yet. Often when writing a book you learn a lot more about a topic than actually gets into the final work. I know that the McCulloch-Pitts neural network is going into Chapter 12, but Chapter 12 needs to have lots of other stuff as well, and I really can't see the neural network getting more than a page or so. What will make the decision very hard is the personage of Walter Pitts: There are multiple bizarre anecdotes about this guy, and two of them involve Bertrand Russell!

Would I have preferred to do all my research at home rather than trudge out to brick-and-mortar libraries? But then I would have missed out on walking around Manhattan on this lovely day! And I always feel so good on Library Day. I feel like I'm really connecting with researchers and writers who have come before me, rather than just typing search terms into Google and clicking through links.

Obviously, I do a lot of research online. Sometimes I find entire published articles that are useful, and sometimes I cough up money to download an article not available elsewhere. In one sense, I'm looking forward to the futuristic age when Google Book Search gets its act together and actually provides a service that resembles a library rather than a heap of disorganized scans.

But I'll also miss the library experience. I like going to the library. It's fun.

* In the early 1970s, when the card catalog consisted of actual cards stored in long drawers in Room 315, the NYPL decided to publish the card catalog, in part so it could be available in other locations, and also to have a record in the event that the cards themselves were destroyed in some tragedy. The Dictionary Catalog of the Research Libraries of the New York Public Library, 1911-1971 was published in 1979. These are known as the "black books" because they have black bindings. The pages of the books are about 14" by 10" and each page reproduces 21 cards, somewhat reduced in size. Each volume has about 450 to 550 pages, and there are a total of — brace yourself — 800 volumes. In Room 315, they now occupy some of the space where the cards once lived.

The black books proved to be useful when the NYPL converted to an electronic catalog system. The library could focus first on getting all the entries from 1972 into the system because the black books could be referenced for anything prior to that.