veryard projects - innovation for demanding change


book review

Bruno Latour, Aramis or the Love of Technology.

Translated by Catherine Porter. Harvard University Press, 1996.

buy a copy

buy a copy

A brilliant, original and stylish book, by a deep and thoughtful writer. Describes and analyses the twists and conflicts of a typical R&D project.

Phenomenon 1: High reuse of case study material

It is often much easier and quicker to illustrate an argument using a well-known case study, than to present a new case study in sufficient detail to support the argument.

As a result, we see the same case studies used by many writers to make many different (and often contradictory) points. An extreme example of this is the Challenger disaster, which has been analysed in countless different ways. In this single example, the failure of the O-Ring is traced to all kinds of flaws: in the engineering process, in the management structure, in the political environment, or in the unconscious psychology of the organization. Other case studies are not so much analysed as paid homage: mention the London Ambulance Service, for example, and the reader is expected to nod wisely and adopt a fixed critical stance.

It is therefore a delight to come across the book under review, which documents and analyses a large yet little-known R&D project in enormous detail. The book is witty, insightful, and is written in the style of a detective novel, following the progress of two academic sociotechnologists as they try to conduct a postmortem on the Aramis project. Who killed Aramis, and why?

Aramis was an attempt to construct a new mode of public transport in Paris, which would combine the privacy and directness of the private car with the efficiency and speed of the metro. The project lasted from 1970 to 1987, absorbed many million francs, produced several impressive prototypes, gained political support from successive administrations, and was suddenly and summarily cancelled. Our two fictional academics are trying to explain on the one hand why the project lasted as long as it did, and on the other hand why it didn't complete. They interview the participants, wade through the files, and come up with a succession of hypotheses, each one modified or undermined by the evidence uncovered in the next chapter. Latour alternates the fictional discussions of the academics with extracts from the real-life material gleaned in his field research, tracing the technical complications, politically motivated half-truths and ever-shifting and competing visions with clarity, intelligence and rigour.

Many writers will be able to find illustrations of their pet theories in the book, although I suspect it will be difficult to come up with good explanations of the phenomena that are not already in the book. To give you a taste of the book, let me briefly mention a few of these phenomena.

Phenomenon 2: 20-20 hindsight

When interviewed retrospectively, many of the participants claim to be able to identify some fatal flaw in Aramis that caused its death, either in the technological concept itself, or in the socio-political environment in which Aramis was to have been implemented. However none of the participants expressed any reservations at the time, except perhaps in the standard small-print phrases that are used to evade liability. Even when the prototypes didn't actually work, or didn't demonstrate what they were supposed to demonstrate, these failures were always explained away in the interest of keeping the funds coming. All the participants apparently colluded in remaining optimistic.

The problem with the 'fatal flaw' type of explanation is that it explains too much. Similar 'fatal flaws' can be often found in projects that did run to completion. In this particular case, technology analysts are especially fortunate; Aramis had a sister project called VAL, which involved many of the same participants at the same time, and shared many of the allegedly 'fatal flaws' that are blamed for the termination of Aramis. VAL was successfully implemented in Lille, and has since been exported to several US cities.

Furthermore, the 'fatal flaw' type of explanation cannot explain why Aramis survived so long. It is certainly not the case that some flaw was initially hidden, and that it was the discovery and proof of such a flaw that triggered the cancellation of the project. Indeed, most of the flaws were known in 1984, when Aramis was given a political and financial commitment to go ahead.

Phenomenon 3: Minor components that take over the show

In the conventional view of technological development, the designer establishes a mapping between a set of demands (usually known as The Requirements) and a set of components or services that collectively satisfy the demands (usually known as The Solution). As is often the case, Aramis turns out to require several components that haven't been invented yet, although engineers are confident that they could develop leading edge technological solutions, given time and budget.

But the design logic of component-based development isn't compatible with the research logic of technology breakthrough. The components that don't yet exist create 'holes' or 'attractors' in the design structure, which have a distorting effect on the project as a whole.

For example, Aramis demands the invention of a new type of electric motor. The variable-reluctance motor is invented, and then this invention starts to be cited as one of the beneficial spin-offs of the project as a whole, although when pressed nobody can think of any other use for this motor. The design of Aramis now revolves around this motor.

Phenomenon 4: Killing by simplification

The development of Aramis was marked by compromise. The initial vision, apparently simple, turned out to contain several conceptually distinct parts. On the one hand, the simultaneous achievement of all these parts was thought to be politically impossible and/or prohibitively expensive. On the other hand, dropping any of these parts was regarded as an unfortunate watering down of the Aramis vision, to the point where the very identity of Aramis was compromised.

In other words, some engineers argued that THIS requirement makes Aramis unimplementable, while other engineers argued that without THIS requirement, there is really no point in continuing with Aramis at all.

Phenomenon 5: Autonomous technology

Like many complex technological projects, Aramis seems to have taken on a life of its own. There is too much at stake for any of the participants to make decisions that conform to textbook rationality. Perhaps this is why Aramis has a momentum, and takes as many years to stop as to start. In the fictional parts of the book, Latour reflects this by giving Aramis a voice, with tongue-in-cheek references to Frankenstein and other monsters. Particularly in these sections, the book hovers on the boundary between sociological analysis and literature.

The book is a marvellous and rich narrative, clear and thought-provoking. Every engineer should recognize with wry pleasure the excitements and disappointments, the debates and distractions and diversions and difficulties of technological development. Strongly recommended.


This review was originally published in Requirenautics Quarterly.


Since writing and publishing this review, a couple of friends (namely David Iggulden and Tom McMaster) have told me that Latour's earlier work is just as good, and I should have started with his Science in Action. This book is indeed also excellent, but rather different in style. Read both - I don't think the sequence matters much.

More recently, I had the privilege to attend a lecture at Brunel University, on April 1st 1998, where Latour offered a controversial, and possibly tongue-in-cheek, analysis of Virtual Reality: "Thought Experiments in Social Science: from the Social Contract to Virtual Society".

STOP PRESS: I had understood that this lecture was to be included in his next book. I have now got a copy of the new book - Pandora's Hope - and it isn't there. Lots of other good stuff though.

Bruno Latour, Aramis or the Love of Technology.

Translated by Catherine Porter. Harvard University Press, 1996.

buy a copybuy a copy
Bruno Latour, Science in Action.

Harvard University Press, 1987.

buy a copybuy a copy
Bruno Latour, Pandora's Hope.
Essays on the Reality of Science Studies.

Harvard University Press, 1999.

buy a copybuy a copy

Contact Details


[Bruno Latour]

[book index]

[technology change management]

veryard projects - innovation for demanding change

Richard Veryard is a technology consultant, based in London.


Review written June 1997.
Technical update on June 15th, 1999.
Copyright © 1997 Richard Veryard