NOTE: This transcription was contributed by Martin P.M. van der Burgt, who has devised a process for producing transcripts automatically. Although its markup is incomplete, we believe it serves a useful purpose by virtue of its searchability and its accessibility to text-reading software. It will be replaced by a fully marked-up version when time permits. —HR
| Copyright Notice | |
| The following manuscript | |
| EWD 539 Mathematics Inc., a private letter from its Chairman : | |
| is held in copyright by Springer-Verlag New York. | |
| The manuscript was published as pages 184�187 of | |
| Edsger W. Dijkstra, Selected Writings on Computing: A Personal Perspective, | |
| Springer-Verlag, 1982. ISBN 0�387�90652�5. | |
| Reproduced with permission from Springer-Verlag New York. | |
| Any further reproduction is strictly prohibited. | |
1st December 1975
Mathematics Inc., a private letter from its Chairman.
Dear ... ,
Yes, indeed, it has been a hectic year! Thank you for your kind feelings. As a matter of fact it started already around Xmas last year, when the rumour reached us that the International Research Development Corporation IRDC was trying to penetrate our market! IRDC is represented by Obfuscate et al., that old clannish sollicitors firm in Oldcaste-upon-Time, which —as luck would have it!— employs a former classmate of mine. I wrote him a letter —full of sugar, you may be sure!— as if I were appealing to him for legal advice. It all worked our beautifully, he even did send us a draft contract, thus providing us with all the information we wanted to have! It was all most reassuring: IRDC is so firmly entangled in legal complications that they are no longer a serious threat. Our monopoly is safe — and in case of problems: we have arranged a secret affair between the nightporter of the Hosanna Building and the second daughter from old Obfuscate’s first marriage, so blackmail is always there as emergency exit.
The whole affair had one nasty consequence: in our moment of panic we felt that we had to do something, and our Proof of the Riemann Hypothesis has been brought out into the field, contrary to the advice of our marketing manager who felt that it still required too much maintenance. And right he was: we can —and do!— burn our stove with the incoming trouble reports! End of March we have transferred fifty mathematicians from Production to Field Support, thus solving two problems at once.
Business being what it was, something had to be done about production, for our stock of unsold theorems was growing beyond the acceptible limits. I have always argued that we should have a reasonable amount of spare theorems in stock, but in March they already occupied nearly two full floors of the Hosanna Building! Besides the transfer of the fifty mathematicians —we have, of course selected the fifty most productive ones— we have returned to our old method of productivity measuring: since February 1974 we measured mathematician productivity by the number of new results obtained per month; we are now back on the more realistic and, after all, also more objective technique of counting the number of lines of proof produced per week. Thanks to those two measures, the stack of unsold theorems, I am happy to say, is slowly shrinking back to normal size.
But for a few little, specialized firms (one in finite geometry and another one in combinatorial logic), Mathematics Inc. has now full control of the mathematical market, a circumstance that is certain to create both political and economic problems. It is not yet an open battle, but the first symptoms of revolt against our dominance become visible for the discerning eye.
For the time being we have nothing to fear, for our greatest allies are and remain the universities, their departments of mathematics, I mean. They should fight us to death, because the more we proceed, the more obsolete they become, and in the end they will be abolished as superfluous. But the technique is so simple! One just sponsors a conference that one calls a “symposium” with only invited university professors as participants. One chooses a nice subject like “The Impact of Mathematics on Society in the Eighties” or “The Role of Mathematical Education in Preparing for the Future” or any other nil-topic. They are so flattered, they come in as an eager flock, proudly carrying their badges home when it is all over. It is pathetic! But also absolutely effective! Did you know that our Differentiation Kit is now used at 378 universities, all over the world? All their alumni will have to subscribe to our “Journal of Kit Differentiation” for the rest of their lives, if they want to remain up to date. The whole movement has now such an impetus that it proceeds without us pushing it anymore; the French have already founded a separate Society for Theoretical Kit Differentiation. It is the same story with our Linear Algebra Kit, our Integration Kit, and our Statistics Kit. It fully absorbs and paralyzes them, leaving the field open for us. Our only obligation is to modify the Kits regularly, that is, to change their appearance slightly, just enough to suggest progress. And really, the universities love them: they always fall for the newest model! They feel themselves superios to the other backward universities and colleges that have not yet converted to Kit Mathematics.
You know that the overall economic, political, and social aspects of this whole venture interest me more than the purely technical issues. But the latter are intriguing too! As soon as Mathematics Inc. grew beyond one hundred employees —can you remember how long ago that was? it seems ages...— it was clear that, no matter what we would tackle, the diversity of our products and manufacturing techniques would create havoc from the organizational point of view. As standardization of products is only possible to a very limited extent —the market place somehow insists on variety— we had to standardize our manufacturing techniques. And we are proud of our IR System for Integrated Reasoning, and I think justly so. In the beginning the IR System was not too successful, because we needed a computer and chose the then fashionable 1033-alpha, a machine for which the MTBF transpired to be of the order of magnitude of twenty minutes. The IR Systems works much better since we have replaced it by the new model 1035-omega, for which via a switch on the console the parity check can be disabled. As soon as we had the new machine, IR made significant progress: the whole IR System now consists of the following languages:
| ASL | Axiom Statement Language |
| LSL | Lemma Statement Language |
| TSL | Theorem Statement Language |
| PSL | Proof Statement Language |
| PVL | Proof Verification Language |
| PRL | Proof Refuting Language (our main debuggin aid) |
| IL | Inference Language. |
Their mutual relation is roughly as follows
| Legenda: | arrows denote causal connection. |
| means that the vertical one controls the horizontal one. |
Actually this is a slight simplification, because it refers only to the high-level languages PVL, FSL, and PRL, while we have the corresponding low-level languages pvl, ps1, and prl as well: they are only needed when the IR-facilities need more efficient exploitation. As you see immediately from the above diagram, ASL, LSL and TSL are purely syntactic languages without any semantic contents, PSL is a language with an ambiguous, nondeterministic syntax and only first-order semantics, while only PVL and PRL have second-order semantics. IL —although we call it Inference Language for the sake of homogeneity— is, of course, no language at all: it is no more than the specification of the interpretation that can supply quasi-semantics for ASL, LSL, and TSL. It is amazing that people have never thought about the coupling of reasoning controlling activities, but once you have got the idea, it is clear that the above scheme provides all the facilities you may ever need. Our IR System —which, of course, is for internal use only— has been very successful. (I have heard rumours that the application of PRL to Riemann Proof will require a 1033-omega-super, extended with a quadruple-length complex arithmetic unit. Some guys are so fanatic that they want to order one, but —thank goodness!— I hold the purse strings, and I think that I can convince them that also in this case we had better stick to the company policy —which, after all, has always been very successful— of leaving the last stage of quality control to those rare customers that think that they ’really need that extra quality.)
Needless to say, I consider the main benefits of the IR System to be psychological and sociological: the presence of the tool has effectuated more homogeneity in the company than regulations could ever have achieved. There was a time that our topologists could not communicate with the number theorists, they lived in different worlds, although they could work on the same floor! But the IR System provides a standard, common universe of discourse, and, again they understand each other. You can believe me or not, but the other day I saw one guy of the Riemann Group and another guy from the Four Colour Project exchange a few IL-cards! I cannot describe to you how happy that observation made me: at that moment I knew that I had founded a living company. Semper floreat et crescat!
Yours ever,
| Edsger W.Dijkstra, Chairman | |
| Mathematics Inc. | |
| Hosanna Building |
PS. Secret! We are, of course, constantly trying to protect our company against the possible consequences of changing attitudes, and we are not blind to the current anti-intellectualistic undercurrents in our society that rouse popular feeling against Reason. We keenly observe the semi-mystic “back-to-nature” movements that want to do away with organization, with power, with shaving, bra’s and socks. They provide an alibi for the second rate college teacher preaching that “truth is dehumanized if it has to be proved” and “true truth is what one feels to be true” etc. In view of this quasi-religious revival, our third Assistant Vice-President is contemplating —I think that that is the right word for it— an Artifical Devotion Department. (Maybe it is only because his wife presently spends a lot of her time “reviving”. Before the AD Department has materialized, she may have lost interest in revivals or he in her....) In the preparatory stage he has designed Canonical Forms for the Seven Capital Sins, and you should see them: they are absolute beauties! It looks very promising, and this could turn into a very interesting daughter of the company. (End of Secret.)
Solicitors
13 Shyster Lane
Oldcastle-upon-Time
| Your Ref: EWD 475 | 29th February 1975. |
Dear Edsger,
I was delighted to hear from you after all these years. My apologies for the delay in replying to your letter, but as you will see, it had to be forwarded to my new address. It was indeed fortunate that you should seek my advice, since my firm represents the International Research Development Corporation. The IRDC is a well-estab1ished non-profit-sharing corporation which seeks to assist creative institutions and individuals to obtain real rewards for their abstract labours. Amongst their clients have been L.D.Vinci, N.Bourbaki, D.Scott, the Department of Machine Intelligence of the University of Edinburgh and the publishers of the Vienna Telephone Directory, to name but six.
To my surprise, it would appear from your letter that Mathematics Inc. has not established a legal department capable of overseeing the protection and exploitation of its creations, and I would therefore suggest that you join forces with IRDC. Contrary to your supposition, recent changes in patent law have enabled Mathematicians to obtain legal protection for their Theorems. IRDC are perhaps the first organisation to take full advantage of this new situation. They have, in conjunction with my firm, started offering their services to Mathematics-based industrial concerns such as yours, and on occasion to individual Mathematicians, although the day of the individual Mathematician is I fear drawing to a close.
I enclose for your information a draft form of the revenue sharing agreement that IRDC would wish you to accept. You will I believe find the provisions most reasonable, although I am afraid they cannot undertake any aspect of the task of theorem maintenance.
Incidentally, although, as you say, it will be necessary at some stage for you to disclose the Proof of your Theorem, it has become the fashion to do this by providing what is called an Implementation, which can be given in such excruciating detail that you need have no fear that your competitors would find it more profitable to understand the proof than to sign a Royalty agreement. Thus, our last patent, for one of the gentlemen named above, contained no fewer than 2973 lemmas, and was based on a novel extension, to three dimensions, of Petri-Net Theory.
I look forward to hearing from you and to joining you for dinner at the Bali. My secretary’s assistant will make the detailed arrangements by telephone.
Yours most sincerely,
| Jonathan Pettifogger. |
E.W.Dijkstra,
Chairman,
Mathematics Inc.
Hosanna Buildings,
Plataanstraat 3
NUENEN
The Netherlands
REVENUE SHARING AGREEMENT:
T H I S A G R E E M E N T is made the day of One thousand nine hundred and seventyfive B E T W E E N EDSGER WYBE DIJKSTRA of plataanstraat 5 Nuenen (hereinafter called ”the Mathematician“ which expression shall include his legal personal representative) of the first part MATHEMATICS INC. (hereinafter called ”the Institution“) of the second part and The INTERNATIONAL RESEARCH DEVELOPMENT CORPORATION (hereinafter called”the Corporation“) of the third part
W H E R E A S by an Assignment of even date herewith and made between the parties hereto the Mathematician and the Institution assigned to the Corporation the full benefit of the theorems and relative Applications for Letters Patent particulars of which are set out in the schedule hereto and it has been agreed that the Corporation shall endeavour to secure the exploitation of the said theorems upon the terms hereinafter contained
NOW IT IS HEREBY AGREED by and between the parties hereto as follows:-
or
IN WITNESS whereof the Mathematician has hereunto set his hand and seal and the Institution and the Corporation have caused their Common Seals to be hereunto affixed the day and year first above written
THE SCHEDULE above referred to
| Application No. | DATE | Title of Theorem | Mathematician. |
| 16766/75 | 9 February 1975 | Riemann Theorem | E.W.Dijkstra |
SIGNED SEALED and DELIVERED
by the above named EDSGER WYBE
DIJKSTRA in the presence of:
THE COMMON SEAL of MATHEMATICS
Inc. was hereunto affixed in the
presence of:
The COMMON SEAL OF THE
INTERNATIONAL RESEARCH DEVELOPMENT
CORPORATION was hereunto affixed and authenticated by:
Authorised by the Corporation
Secretary
Transcribed by Martin P.M. van der Burgt
Last revision 2015-01-08 .