A recent programming text (Coding Course, S Lee) introduces Python, a high-level language launched thirty years back to support human effort in business, science, and education. Decades before Python, computers were instructed via a series of zeros and ones; still important today, this machine language determines how computers process data. Python is easier to understand because its jargon is more akin to everyday English, for instance, Total_stock = Total_stock + s_change, where the programmer has named two variables Total_stock and s_change in an instruction that adds the content of two storage areas and puts the result into the area named Total_stock. So, if 18 items arrived by post and Total_stock was only 6, Total_stock would become 24. All quite logical. The difference between noticeable events in the real world (like the arrival and departure of stock) and the unreflective state of high-tech devices is called system lag, most noticeable late December when adjusting the shower temperature on a cold morning gives a meaningful tardy response.
Now, Python coding may be just forty lines in a GCSE homework assignment, or a longer piece to check the progress of some NASA project. In both scenarios the computing task involves a well-defined problem; acting as problem-solver, the programmer must design logic which achieves whatever has been specified in the problem definition. If a program should crash – unable to handle a mix of numbers and letters – there will a delay while the programming team unravels the logic and changes a couple of instructions. When the original author has departed for greener programming pastures, the delay may stretch beyond an hour or two. Nationwide, more than a decade of regrettable anomalies in Post Office accounting software have ruined the lives of some employees.
While good programming delivers robust logic which meets expectations, social groups lack neutral settings where designed logic can be easily uploaded. Schools, aged factory premises, and modern hospitals have value-laden histories: each organisation evolves from an obscure trail of unique events. The problem of improving the performance of a corporate sales team or college A Level history team seems more difficult to define than questions resolved by Python. Can the sales team be investigated without considering other departments – isn’t the performance of internal sales influenced by shop floor production, or those beavering away in despatch? At best, sales personnel will have been involved in adapting a software package to support their administrative duties – but were they? Likewise, in a college setting, should tutors rank as the main driving force for attainment, or the local secondary schools which diligently prepare college attendees? And since exam questions change every year, are some final exams tougher than others? Regarding both settings, is IQ merely an intriguing topic for academic psychologists – or a subtle influencing factor across education and the world of work? For unique social settings invariably richer than else-if statements, an alternative type of logic might help unravel the social world.
At the outset of a thirty-year research programme based in the Department of Systems Engineering at Lancaster University, Professor Peter Checkland picked up the gauntlet of devising a ‘systems approach’ suited to the improvement of hazy, real-world problems. Acting as itinerant consultants on behalf of fee-paying clients, Checkland’s small teams found it vital to drop a fashionable assumption that the modern world was composed of systems amenable to expert intervention, often via computer analysts. During hundreds of studies into obscure scenarios (where clients admitted unease over what to do next), a novel approach emerged – then enthusiastically applied and refined in the ever-changing milieu of real organisations. Christened Soft Systems Methodology (SSM), the most recent version of Checkland’s approach has four related stages: first, enquiry into the hazy setting, second, the construction of models appropriate to issues, third, debate over models with concerned participants, and fourth, implementation of any changes agreed because of third stage parley. Essentially, this recursive methodology arranges a process of learning and debate, repeating the four-stage cycle as often as required by those involved. Projects tackled by Lancaster include giving assistance to a perplexed independent consultant, reporting on project management (being almost absent) in a major aeronautical venture, and an exploratory exercise which appraised work done on a clinical information system at Huddersfield Royal Infirmary. The latter brought together information systems experts with four hospital consultants, yet again helping concerned participants learn from the experience of using SSM in an unclear situation.
In contrast to the specialist computer models of ‘how to’ in Python and other related languages, models in SSM are conjectures which in logic pursue a notional mission by means of English phrases devised within a client’s unique scenario. Debate in stage three assesses not just whether a model is rational but – more importantly – whether it seems relevant to participants’ diverse views. Each model displays seven or so linked activities, for instance, ‘estimate the region’s number of wind turbines’ or ‘define criteria for selecting the committee’s secretary’, both user-friendly phrases are not mere ideology but could be put into action. A popular way of adopting stage three is to ask questions of the activities expressed in each proposed model: Is this activity conducted within the organisation? How is it done? How is its performance measured? Who is in charge? and What would count as improvement? Quite reasonably, programmers’ coding is never submitted to such scrutiny; for wider and ever-changing organisational issues, this comparison between abstract models and the real world brings structure to open debate over potential change.
Checkland’s flexible research principles are not in competition with Python, Microsoft Office or any customary ways of working. Rather, the methodology’s guidance offers a participative style of enquiry for ongoing problems which are proving difficult to define. Working outside of autocracies, SSM is a conscious and explicit way of managing enquiry across diverse social groups.
Neil Richardson
Kirkheaton
