# Why was the original typesetting of Number Fields by Marcus so horrible?

Does anyone here know what technology or instrument was used to typeset the first edition of the well-received textbook Number Fields by Daniel A. Marcus? I ask because the original edition looked like it was typeset using a toaster (forgive my exaggeration). A preview of the first edition can be viewed on Google Books. It was published in 1995 by Springer, whereas I have read papers from much earlier where the typesetting was pretty good (for example, this paper of Lehmer from 1932), so I can't fathom why the typesetting was so bad. It looks like it was written on a typewriter, though given the usage of notation from algebraic number theory, it probably was not an actual typewriter.

(Thankfully, a LaTeX-ed second edition was finally published by Springer in 2018.)

The book came out in 1977, not 1995; look at the year at the end of the Foreword that Marcus wrote. My copy has a copyright from 1977 too. LaTex or even TeX was not an option in 1977 since that was the year before the original version of TeX was released. The typesetting in the original version of Marcus was clearly done on a typewriter.

I agree it is unusual for a professionally published book from the 1970s not to be typeset better, but if you look at Springer Lecture Notes in Mathematics from that period you will find volumes that were prepared on typewriters too. Among math books, Serre’s Lie Algebras and Lie Groups (1965), Fulton’s Algebraic Curves (1969), the Proc. Symp. Pure Math volume on Mathematical Developments from the Hilbert Problems (1974), and Mumford’s Red Book (1976) were all prepared on typewriters. So I’d say Marcus was in good company. Around 10 years after "Number Fields" first came out, Cambridge Univ. Press published "Local Fields" by Cassels in 1986 and it was quite clearly prepared on a typewriter.

That there are math symbols (Greek letters, doubled Z and R for $$\mathbb Z$$ and $$\mathbb R$$, etc.) does not mean it could not be done on a typewriter. Typewriters had special mechanisms to type a range of technical symbols. Articles were submitted to journals back then as typewriter copy.

Look here for a discussion of mathematical typing before LaTeX.

• I agree. This was what you could do with a typewriter in 1977. Since typesetting (which could produce much better appearing copy) was expensive, this method was often used in those days. A large math department would have among its secretarial staff, those who could produce this kind of result. – Gerald Edgar Jun 18 at 11:45
• And, long ago, the Springer Lecture Notes explicitly said that they were devoted to inexpensive, rapid availability. In those days, this meant literal typewriters (of some kind), not typesetting. Typesetting would increase cost, and delay things substantially (no internet to communicate about edits...). I'm sure the other people who avoided typesetting their books wanted to achieve similar. I think the first edition of Spivak's "Differential Geometry" was typewriter-written, too. – paul garrett Jun 18 at 17:42
• @paulgarrett: "no internet to communicate about edits" -- minor nitpick, really, but the first email dates back to 1971. ;-) – DevSolar Jun 22 at 3:33
• @DevSolar even if email existed in principle in 1971, I don't think it was actually being used in practice on a wide scale among university faculty until the late 1980s, and I suspect that including companies like Springer would push the practical date farther into the future like the early 1990s. – KCd Jun 22 at 4:20