Previous column

Next article

---

On an open issue of programming language phonetics Bertrand Meyer, ETH Zürich, Switzerland

COLUMN PDF Version

Abstract

For decades or more, a vexing question has plagued vocally-inclined mathematicians wishing to communicate in accepted variants of the English language: how properly to pronounce the word “tuple”. The issue has recently taken on a new urgency with the introduction into Eiffel of a widely used notion of TUPLE type. Yet progress towards a solution has been painfully slow. We propose a novel approach that we hope will answer the question for generations to come.

1 THE PROBLEM

When considering the wealth of literature, conferences and curricula devoted to matters of programming language syntax and semantics, one cannot fail to be struck by the paucity of quality material available on programming language phonetics. The consequences on the state of the software industry are there for everyone to contemplate. How should software engineers be expected to convey the correct meaning of a program to a computer if they cannot even read it aloud to a fellow human in an unambiguously understandable form? Imagine a C programmer shouting, across the cubicle lane, a typical informative text such as “Ampersand Star Star Bracket Curly Curly Plus Plus”; what if the listener was trained as a logician, to whom “Curly” might sound as “Curry”; or as a quantum physicist, to whom a Bracket is a very special kind of operator? Or as an astronomer? In the absence of reliable statistics — another telltale sign of the dismal state of research in this neglected discipline — one can only speculate how many software failures follow from such cases of programming language mispronounciation.

Nowhere is the situation more damaging than for tuples, an important programming language concept now widely used in Eiffel, and borrowed from mathematics in both concepts and terminology. Experts disagree, sometimes acrimoniously, on the proper rendering of the “u”: as in Bunny, or as in Dude? Both forms, indeed, are commonly heard from seasoned tuple practitioners, their distribution straddling the usual linguistic divisions such as British vs. American.

The present work proposes an original, highly poetic solution to this critical issue which we hope will both address the practical concern of industry and spur a fresh growth of innovative research in programming language phonetics.

2 THE ANSWER

ACKNOWLEDGEMENTS

This work arose from an idle doodle scribbled while drinking Snapple at some befuddled (or bedazzled) moment of the Eiffel ECMA Eiffel standards committee’s discussion of TUPLE syntax and semantics, Orinda, California, 23 February 2003. The author is greatly indebted to the other committee members for the fascinating diversity of their speech habits. The final version of this article will appear as [Meyer 03].

REFERENCES

[Meyer 03] Bertrand Meyer: "Men Are From Tooples, Women are from Tupples: Towards a Poetic Ontology of the Politics of Gender in Object-Oriented Phonetics", in TOOPLE/TUPPLE (Tacky Object-Oriented Poetry from Lithuania, Estonia, Tampere, Uppsala, Petersburg, Poland, Latvia Etcetera), Baltic Literary Object-Oriented Proceedings, Edited and Revised (BLOOPER), Riga, scheduled to appear April 1st, 2003.

About the author

Bertrand Meyer is a consultant specializing in pronunciation issues for object-oriented concepts. Rates available upon request (specify language).