High Leverage Interventions

The implications of David Lane’s interesting and provocative paper have by no means been exhausted by the commissioned commentaries by Washburn and by Sodhi and Tang.

I have a number of minor issues with some of the things they say, or don’t say, about Blackett, one of Lane’s 3 trail-blazers, which I will append at the end of these remarks. But the main point I want to make relates to the analyst/client distinction so usefully highlighted by Washburn. He sees the gulf of mutual comprehension and trust between them as a significant impediment to ‘high leverage’ OR interventions of the kind Lane wrote about. In effect Washburn then says that we clearly cannot teach OR students to be context-aware, since the contexts they may work in are so various; so perhaps we should instead try to teach budding executives to understand the basics of OR/MS.

There is a third alternative – to make the production of analysis-based decisions a joint enterprise between clients and analysts. Each partner can then contribute from their own particular knowledge base and expertise. There is a historical precedent for this: the very close collaborative relationship they achieved has been credited as one of the key factors in the success of the World War 2 OR in Britain. Indeed Solly (later Lord) Zuckerman went so far as to describe OR as ‘the bastard born out of a secret liaison of scientists and airmen’ (Zuckerman, 1964, p 287). Such prolonged intimacy may not normally be possible away from the pressure cooker of wartime circumstances. However we can perhaps approximate it for limited periods.

This is indeed one way of describing what happens in a well-run ‘soft OR’ (or Problem Structuring Methods) workshop. The mutual engagement of client and executive is promoted by the transparent, model-based forms of analysis which are deployed, and by the participative process. These enable the development of a mutual understanding and trust which give the client/executive confidence both in her analytic helper and in the results that have been achieved.

In purely technical problems, of course, such a relationship is not required – but high leverage interventions are never purely technical.

Sodhi and Tang accept the relevance of ‘Soft OR’ to the sort of higher profile OR interventions that Lane is arguing for. But they fear that even mentioning it “generates a rather visceral and negative reaction” not least among journal editors. This journal is not excluded – see OR/MS Today for an article, a letter to the editor, and an editorial on the subject (Mingers 2009; Ackermann et al 2009; Horner 2009). Sodhi and Tang appear to think that Lane steered clear of this issue. Not really – there is an extensive treatment, though phrases such as “modelling tools and approaches appropriate for more sociotechnical situations” are used in place of ‘soft OR’. How will the OR community make progress in its professional deliberations unless members feel they can speak and write openly?

Ah yes, those detailed issues:
i) Washburn states that Morse and Kimball’s historic text (1951) was “the first mention of the dichotomy [between analyst and client] that haunts our profession”. In fact this distinction is integral to the very first OR document – Blackett’s (1941) ‘Scientists at the operational level’ written in the heat of OR’s genesis.
ii) Sodhi and Tang say that “Blackett has some claim to be the father of OR”, but go on to equate his claim to that of Babbage in the 19th century. The significant difference is that Blackett set up and led a series of groups explicitly dedicated to the practice of OR across the three armed services, and was responsible for a whole series of projects triumphantly brought off. Not only that, the current professional practice of OR derives directly and continuously from the work that Blackett (and other contemporaneous scientists) carried out. Some claim!

Ackermann , F. and 46 others (2009) ‘The case for soft OR’, OR/MS Today, April 2009, 20.
Blackett, P.M.S. (1941) Scientists at the Operational Level, Admiralty, London.
Horner, P. (2009) ‘Hard sell for soft OR’, OR/MS Today, April 2009, 4.
Mingers, J. (2009) ‘Taming hard problems with soft OR’, OR/MS Today, April 2009, 48-53.
Zuckerman, S. (1964). ‘In the beginning—and later’, Opl Res Q 15: 287–292.