I'm interested in relationship between two very important Computer Typography systems Troff and TeX. In particular, I would like to find any statements made by Donald Knuth about the influence (or not) of Troff on the early development of TeX78. Could anyone point me to a suitable resource?

I have tried hard to find any records of remarks made by Don related to Troff and its influence on the creation of TeX. I was also preparing to ask him that question personally year and a half ago in San Francisco. Unfortunately, I saw that dramatical iTeX(to be read text) announcement only on video. Since there are many people here who know Don personally, has anybody bother to ask him that question privately?

I am bit surprised to find that this question has not been asked earlier as I recognize many names on this portal who made significant contributions to both systems and who have used computers long before Microsoft "invented Windows". It is a bit inspired by the talk of Brian Reid, the creator of another legendary but now forgotten typesetting system Scribe, about his disillusion with the business of computer typography. I had true privilege attending that talk many years ago. Unlike TeX, it seems to me that Scribe was a legit attempt to take computer typesetting in a "radically" different direction than Troff.

As somebody who essentially grow up using TeX as THE computer typographic system and only learning Troff as a part of general Unix folklore, it would seem to me more logical that Don tried 1978 to fix what was wrong with Troff and continue on his work on Art of Computer Programming instead of creating entirely new typographic system from the scratch. Although proprietary, Troff definitely was not closed source (at least for somebody at Stanford) and it is very formidable general purpose typesetting system. I know enough (I should say little) about both systems and some of their major internal differences to appreciate TeX contributions to the world of Computer Typography. On the another hand, the idea of pre-processors utilized by Troff is very close to my heart being a Unix addict even though \special is indeed very special command in TeX. I like the fact that Troff is tiny comparing to TeX even that one that I use mandoc (OpenBSD user here) and mdoc macros is not at this point general purpose typesetting system (at least not yet usable for Mathematics).

I am really curious to get some answers on this one. Maybe people who abandoned Music TeX in favor of developing LilyPond can also shed some light.

  • 2
    on pp.639-640 of "Digital Typography", is the answer to a question from a q&a session with the dutch tex group. Andries Lenstra asked "Why didn't you start from troff? It was completely inappropriate?" i believe the transcript of the session was also published in tugboat, but i haven't time just now to look for it. Dec 27, 2011 at 18:28
  • 2
    Interesting though this might be, it doesn't quite fit the normal "Question and Answer" format of this site which is for specific answers to focussed questions. In particular, the only sentence that ends with a question mark seems not to be the actual question. May I suggest that you rewrite your question to something more suitable for the format. Perhaps something along the lines of "I'm interested in ... and would like to know where I can find out more information about the influence (or not) of Troff on the development of TeX. Can anyone point me to a suitable resource?" Dec 27, 2011 at 20:05
  • 1
    Knuth's very first TEXDR.AFT (May 1977) references the system of Kernighan and Cherry (eqn, based on troff) which was published 2 years earlier in CACM (March 1975). Based on my understanding, from Knuth's point of view, there had been all these attempts at making computer output more pretty, but they were nowhere close to books (he had high standards, such as Addison-Wesley's mathematical books typeset by Hans Wolf next door). It was only when he saw Patrick Winston's digitally typeset AI book in Feb 1977 that he made any connection between computers and something worthy of books. Oct 8, 2016 at 3:08
  • 1
    Another point: although associations with Linux/BSD/open-source today make Unix seem like a common denominator available to everyone, in 1977 Unix was expensive and/or exclusive: it was available to you only if you were at certain select institutions. The world of Unix and Bell Labs did not have a direct bearing on Knuth using the computers in Stanford and SAIL. So although he was aware of published work such as eqn and based some of TeX's syntax (especially math syntax) on it (the syntax is even closer in those draft versions), troff probably wasn't something he used (though PUB was). Oct 8, 2016 at 3:19
  • 1
    I remembered this question and asked Knuth after his Christmas Lecture Thursday (while standing in line of people who wanted books signed/photos taken) about influence of earlier systems on TeX. He mentioned PUB and also "the Bell Labs system" as giving him existence proofs. He had used PUB to typeset errata to TAOCP books. But, he said, the best way to trace TeX's history was to look at his memos published in Digital Typography. He went into how people told him he should have used more of a nested structure like PUB, but his mind doesn't naturally close parentheses (like "end comment") etc. Dec 13, 2016 at 18:50

3 Answers 3


On 13 March 1996, Don Knuth met with members of the Dutch TeX group (NTG) for a Q&A session. During this session several questions came up that alluded to troff. The transcript of the entire Q&A session appears in TUGboat 17:4, pp.342-355, with the immediately relevant material starting on p.348. But the underlying philosophy is laid out in the answer to the question from Frans Goddijn that starts on p.347.

After some discussion that mentioned troff, Andries Lenstra asked "Why didn't you start from troff? It was completely inappropriate?" This perhaps doesn't completely answer your question, but the short answer is that troff at that point was layer on layer of patches, and Knuth felt that another patch wasn't the answer -- time to start all over. (After all, as he pointed out a little later, he scrapped TeX78 when he decided that he had found a better approach, and wanted a clean implementation.) It's not so hard to understand when you realize that he's a perfectionist.

Quite a few of Knuth's Q&A sessions were transcribed in TUGboat; to see all of them, look for the heading Knuth interviews and Q&A sessions in the list of contents by keyword.

  • 2
    One should note, however, that TeX syntax for math is heavily influenced by that of troff (or eqn, if one prefers).
    – egreg
    Dec 27, 2011 at 20:53
  • 1
    Barbara, I know that you have known Don personally for many, many years.Has he ever spoken about Troff apart of that 96 q&a and its creators (unfortunately Joe Ossana was already diseased)? I would imagine that he was instrumental in brining Brian Reid to Stanford. Apparently, he was interested about other systems too, at least in the same fashion we read other people's publications? Dec 27, 2011 at 21:31
  • @Predrag -- i haven't talked with Don about troff or eqn, but there are other references in "Digital Typography". neither Scribe nor Brian Reid are mentioned in "DT" (they're not in the index, and the index is very complete). they are mentioned in the TeXbook (pp.371 and 404), but the references are in an end-of-chapter quote and a sample letter, so those references are more "by the way" than substantive. the letter places Reid in electrical engineering, and Don was in comp. sci., so it's not clear to me that he was instrumental; i don't know. but interested in other systems? yes! Dec 27, 2011 at 22:54

I recall reading something that Don said that he was fortunate being in Stanford where he was free to start over from scratch, whereas if he had been in an environment like Bell labs he would have been compelled to use troff, because it was invented there.

Unfortunately, I can't find where I read this. Maybe it was part of the interview series published in his lastest book Companion to the Papers of Donald Knuth

  • 10
    I'm pleased to welcome in TeX.StackExchange a well known member of the TeX community. Have a nice stay, we all look forward for your valuable contributions.
    – egreg
    Mar 9, 2012 at 10:55
  • 1
    hi, @Ulrik -- the reference is in "digital typography", p.639. it's in answer to a question from andries lenstra raised at a january 1996 q&a session with members of ntg (the dutch group), and can also be found in the ntg maps, p.43, as well as in tugboat on p.349. (the maps version has pictures, as does the one in digital typography; we didn't get them for tugboat.) Mar 9, 2012 at 16:30

I am resurrecting this thread to post links to two interesting documents about typography and computer typography in particular. The first one is the link to PowerPoint slides of the legendary lecture by Brian Reid (at that time he was working for Digital Equipment Corporation after being denied tenure at Stanford). It is reflection on the history of computer typography through his personal journey as one of the main actors. Here is my personal copy. The second one is the link which I already pointed in the thread about use of the LaTeX in industry. TeX & troff -- The Red-headed Stepchildren of Typography? I hope that people will enjoy reading...

Another outstanding Troff specific reading. Must for any Unix funs out there.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.