Interview with Dr Stephen Castell

Dr Stephen Castell was born in Cairo in 1946. His father was in the RAF and the family returned to the UK when Stephen was a baby. Stephen has an older sister. The family moved to different RAF bases around the UK and the world, including those in Scotland, Northern Ireland, Malta and Germany. As a result of moving around, Stephen went to eight different primary schools and three different secondary schools. Stephen says: “I’ve always enjoyed school.  In fact, I enjoyed the fact that we moved every two and a half to three years and I had the benefit of encountering new places and cultures, and the challenge of making new friends, each time. … My parents always regarded education as being the most important thing, which is a lesson which I have learnt and it’s a lesson which we all should learn.”

Having sat his Eleven Plus, Stephen began his secondary education at Magnus Grammar School, Newark-on-Trent, Nottinghamshire, continued at Prince Rupert School, Wilhelmshaven, Germany, and finally went to Hamond’s Grammar School in Swaffham in Norfolk, where he took Pure and Applied Mathematics and Physics A levels.

After his A levels, Stephen went to the University of London, where he studied for his first degree in Mathematics and Physics, and Psychology.  As his degree progressed, Stephen had to drop Psychology but says: “I’m still very pleased that I did it. I had an affinity for it, and it raised all kinds of things which actually come into IT as well; the whole of the ‘meaning of meaning’, epistemology, philosophy.  All IT people should be philosophers as well.” While he was in London, Stephen learned in 1965 to program in FORTRAN on an IBM machine which used punch cards.

Stephen graduated with a First-Class Degree, scoring the highest grades in Physics that the University of London External Examiners had ever seen. He then moved to the University of Nottingham to study for his MSc in Mathematics, Computer Science and Fluid Mechanics, followed in 1972 by an Award-Winning PhD in Mathematics, ‘Certain Transformations in Gas- and Magnatogas-Dynamics’.  Stephen says: “If anyone were to do a study on Mathematics PhD theses in Nottingham, they’d probably find that mine was the first one that actually included computer programs and graph-plotter printouts in it.”

Stephen’s second encounter with a computer was at Nottingham which had an English Electric KDF9.  Stephen explains: “We used to program on punch tape, and then submit it into a batch run, and if we were lucky it would get done the same day, depending on what the loading was, and then get a huge stack of printout the next day. We learnt proper Computer Science, including Phrase Structure Grammars, and how to write a Compiler.” ALGOL 60 was the first programming language in which Stephen became proficient.

In his fictional article, ‘Ye Nom De Das Geld’, published in the December 1971 issue of GONG, the student magazine of the University of Nottingham, Stephen introduced his conceit of a ‘Post-Purse Paradise’:  “Brothers and sisters, I welcome you to the post-purse paradise.  … Geld is in heaven, all’s well with the world. … Cromstock and I first mooted the possibility of an Economic Reformation taking place in Britain in The Journal Of Comparative Economics during … 1969.  … to put into practice … the tenets of the Quasicurrency Theory which I had been formulating over the preceding twenty-five years.  …”.

In 1971, two years into his PhD, Stephen took a role in industrial research at Chalfont Park in Buckinghamshire (a location often used in film sets, eg Bond movies) with the British Aluminium Company (BACo), which was part of Tube Investments; a holding company for a lot of the metals industry of the UK.

Stephen worked on mathematical modelling, using numerical analysis and computational methods, to solve process engineering problems in the aluminium industry using an ICL 4120. The Government had installed a brand-new aluminium smelter at Invergordon which smelted aluminium by the Hall–Héroult process.  Stephen says: “I was working specifically on modelling the voltage drop inside the carbon baked anodes. The equation at the time was that every millivolt that you saved was worth about £50,000 a year in electricity costs. By the time I had left about a year and a half later, we had actually done theoretical calculations on the computer and proposed engineering modifications which saved them about a 50 millivolts drop, £250,000 a year; a lot of money back in 1971.”

Afterwards, Stephen heard BACo’s commercial procurement arm had managed to do a better deal on buying the electricity from Dounreay and had cut their electricity bill by about a million a year.  He says: “Who’s really important, the scientists or the commercial people?  I thought that’s an interesting lesson. There’s an awful lot of technology and science and engineering you could do, but, if it’s about serving a commercial interest, where should you put the effort?  Isn’t it more worthwhile to have better trained procurement professionals than scientists?  An interesting question.”

 Stephen also worked on a project involving practical magnetogasdynamics, a highly complex and esoteric field which was at the heart of his PhD. He explains: “I actually tried to model the way that the molten aluminium conducting fluid swirled around the smelting tank.”

In a correspondence with Henry (Solomon) Lipson CBE FRS (11 March 1910 – 26 April 1991), Professor of Physics, University of Manchester Institute of Science and Technology, 1954–77 (co-inventor of the aid to calculation, Beevers–Lipson strips, which were widely used in the days before computers), published during October-December, 1973, in the monthly magazine of the Institute of Physics, Physics Bulletin, Stephen said “… the computer is the ultimate simulator; … and one day the 20 year old computer will perform better in this field than we 10000 year old humans with our crude heuristics and intuition.  One day, the machine will churn out new ‘laws of nature’ two-a-program; we should all look forward to that day in hope”.  Stephen has always been rather proud of Professor Lipson’s final retort in Physics Bulletin, December 1973, Letters, p750:  “I read the letter from S P Castell (Physics Bulletin October p627), opposing my views on the computer, at first with amusement and then with rising incredulity. But then a possible answer occurred to me. Is S P Castell the pseudonym for a digital computer?”.

In 1973 Stephen moved from BACo to RHP in Chelmsford, where a new Bearing Research Centre had been created. Stephen became Senior Mathematician, reporting to the Director of Research. He was involved with tribology and fluid mechanics. He says: “This was good from a professional interest, as well as being in a company that was looking like it was trying to be a leader in advanced rolling bearing design and manufacture.  We did some quite interesting research about noise reduction, vibration, and lubrication.”

Whilst at RHP Stephen says he had the honour of becoming and serving as the Honorary Secretary of Chelmsford Engineering Society, one of the oldest such organisations in the country.

Since the end of the First World War, engineers from local companies (Marconi, Hoffmann, Crompton-Parkinson, Christy Norris etc – then all major players) would gather to discuss their interests and relevant technical matters. This group was known as the Chelmsford Junior Society of Engineers (CJSE).  Apprentices from Crompton-Parkinson formally joined forces with Hoffmann and Marconi, and on Wednesday 13 October 1920, the Chelmsford Engineering Society (CES) was launched.

Now called the Chelmsford Science and Engineering Society (CSES), it therefore celebrates its 100th anniversary in 2020, and to mark the occasion the Society, in partnership with Anglia Ruskin University, successfully bid to host the British Science Festival, bringing this international festival to Essex for the first time in its 189-year history. Indeed, Essex 2020.com is in 2020 specifically celebrating through Science, Technology, Engineering, Arts and Maths (STEAM) Chelmsford’s pioneering past.  As noted on the CSES website (https://www.chelmsfordses.org.uk/about/sponsors), Stephen,  becoming its Honorary Secretary back in 1973, took over from Eric Payne who had personally known and worked with Signor Guglielmo Marconi himself, the Nobel Prize-winning physicist and inventor credited with the ground breaking work necessary for all future radio technology.

Unfortunately, RHP itself did not fare well against competition from the Japanese and Stephen decided that he could not see much of a future for him in mathematics and scientific applied research in industry.

Having decided to move on from RHP, Stephen started to look at consultancies and, with offers from many of the big names, he opted to work for Touche Ross who had recently started their Management Consultancy. He says he chose them because: “It was smaller, cosier. I liked the people I met there, and they were just starting out.  I often like new things that are starting up. The fact that it was part of one of the top five accountancy firms, as it was then, meant that there were all sorts of opportunities to learn all about the commercial side of financial management and accountancy applications.  Plus, they were the leaders in computer audit at the time.”

One of the first things that Stephen did was an intensive IBM COBOL course. He adds: “I acquired a reputation for being able to spot COBOL bugs even before you put them through the compiler.”

After his course, Stephen became a Project Manager and among many other projects worked on an online system in the London banking services sector for an international merchant bank called Bremar Holdings Limited.

He explains: “I was thrown into managing a project to put in what was then quite an unusual thing, which was an online system, through dial-up modems. …You could actually fill in a form on the screen, hit a button, and it would scrape all the data and send that all down as a record, rather than having to do each one at a time and get a response back. This was quite revolutionary then.”

Touche Ross also held the contract for auditing the Post Office which Stephen became involved in. He says: “How do you value the debt, the liability, that is represented by all the stamps that people have bought but are sitting in their wallets, in their pockets, in their drawers?  You’ve got to make an analysis, an estimate, because it’s like a prepayment (as for an insurance premium).  So, we used to come up with these amazing models to try and value the multimillions of postage stamps that were sitting out there not yet stuck on the letter.”

For the first time in his career, Stephen was managing people. He says of his style: “My management principle is that, ‘he manages best who manages himself’.  I really don’t have time for people who can’t manage themselves. … To be fair, I never had to manage a large department in a company where you could have a whole range of standards of skill, motivations and training.  It’s usually been at the fairly rarefied end of people who are top consultants in the IT world; those are actually the teams I am good at managing.  But mainly because they’re pretty good at managing themselves.”

He adds: “The CASTELL motto, my consultancy motto, is: the Sharpest Sword is Forged in the Fiercest Flame.”

Stephen tells a tale of how he was tempted to write a book about management after his experience of working as Business Development Consultant for the BBC, successfully building from scratch, in 1986-1992, its world-leading multimillion commercial data broadcasting venture BBC Datacast. After completing his initial appraisal of the business and submitting a thorough review, he had coffee with the head of BBC Enterprises who told him that they appreciated his insights, creativity and rigour, but were not in a position to implement the various changes suggested and instead they would just “muddle through”. Stephen explains: “I’ve heard other people say this about the style of British Management: ‘We’ll muddle through’ … All those world wars, creating the Biggest Empire, for all sorts of reasons – it always seems to have been done on a ‘muddle through’ basis and I thought that’s it, I’m going to write my wonderful, world-winning, millionaire management training book, to be called The Muddle Through Manager, because actually, that’s probably a more practically achievable and successful methodology than anything else.”

Stephen moved in 1975 to be Group Management Services Manager (CIO) for Bremar Holdings Limited, principally involved in further designing, developing and managing Bremar’s real-time Eurocredit and Eurodollar Trading Systems, and in Corporate Finance and Venture Capital.  He also developed and coded a ‘banking paper’ bid-and-offer non-linear programming algorithm, running on early desk-top programmable calculators, for use by Bremar’s traders.  This had essentially the same functionality as, and pre-dating, the 1997 Nobel Prizewinning Black–Scholes Model, assisting traders determine the fair bargain price for a call or a put option based on variables such as volatility, type of option, underlying paper price, time (including timing ‘rests’ of interest payments), strike price, and risk-free forward rates.  He was also initiator and manager, as Tournament Founder and Director, of the sponsorship and execution of the London Bremar Cup 1977, the first Colgate Series Women’s Indoor International Tennis Tournament. Stephen personally ideated and arranged for Virginia Wade, legendary Wimbledon Championship Winner in that Queen’s Silver Jubilee Year, to be presented with a unique commemorative trophy, from Erwin Brecher, Bremar’s Founder and Chairman, by Angela Rippon, the world’s first female National Prime-Time TV News Reader.

In 1977 Stephen was co-founder with Neil Maybury, solicitor, of Infolex, the UK’s first commercially-available computer-assisted legal information retrieval service for practising lawyers, hosted on BT’s world-leading Prestel public videotex service, the forerunner to the internet.  This provided a proprietary uniquely-indexed Case LAw Report Updating Service database, CLARUS, which became internationally well-known and much-celebrated in the professional IT and Law community worldwide.  Infolex was bought-out in 1984 by Wolters Kluwer the large international publishing corporation.

In 1978, Stephen formed CASTELL Consulting with the strapline ‘management and financial consultants in information technology’. In the same year, two other consultancies launched; the Butler Cox company and PActel, which also used the then neologism ‘information technology’ in their descriptions.

 In the run up to forming his consultancy and with the recently launched microprocessor technology taking off, Stephen set up a company called Microcomputerland Limited with Paul O’Grady and Brian Reynolds who had one of the first development kits for the Intel 4040 microprocessor. The plan was the then revolutionary idea of a shop selling computers. However, the business was overtaken when Paul and Brian formed Micro Focus to develop their CIS COBOL, and they became too busy to worry about having a shop.

In 1983, Stephen wrote an introductory book for computers called ‘Computer Bluff’, illustrated by Geoffrey ‘Jeep’ Planer. The book became a bestseller of its type, selling about 20,000 copies (1983, Quartermaine House, ISBN 0 905898 15 X: “The Which Computer book for people who know nothing about computers … and would like to have left it that way”).

Stephen has also made a uniquely significant contribution to the mathematics of special large integers.  As noted on page 4 of Computer Bluff, he is acknowledged as the ‘presenter of the most useless fact discovered by a computer’ (Computer Bulletin), being his discovery that “there is only one digit 5 in factorial 82 (radix 10)”.  This discovery was first announced in his 1973 paper, published in Eureka (the Journal of the Archimedeans, the Cambridge University Mathematical Society), ‘On the Distribution of Decimal Digits in n!’, Eureka, 36, 45-47.  A further discovery, that “There is no digit B (i.e. ’11’ in decimal) in factorial 75 (base 16), a 160 digit number”, was announced as a result of trans-border collaborative research, in the 2014 follow-on paper, ‘n!–Forty Years On’, Eureka, 63, 40-41, Dr Stephen Castell, and Ms Forough Khaleghpour (of Sheykh Bahaee University, Iran).

In the early disruptive days of microcomputers, Stephen’s call for an Action Group on Information Technology, AGIT, was extensively published and promulgated; with wider pleas from the software and electronics community, Kenneth Baker MP was appointed as the first Minister for IT, securing a UK position in the PC and Internet Age.  In 2017 he similarly proposed an Action Group on Robot Integration and Control, AGRIC, with a call to Mrs Theresa May, then the UK PM, to appoint a Minister for AI to address the fundamentally new societal and regulatory challenges, and, equally, seize for the UK the new opportunities, arising from this rapidly evolving ‘Intelligent Machine Species’.

Stephen’s initial CASTELL client was Harvard Securities Limited, well-known Licensed Dealers in Securities. He explains: “I was on their letterhead as their advising consultant to go out and find new start-up, new technology companies.  For the next five years I must have looked at 400 to 500 new companies and we floated about six.”  Among the six companies were Grease Eaters, Micromite and Bleasdale Computers, one of the first Unix computer companies in the country.  As a result Stephen became known in the 1980s as a high-profile founder of the UK’s Over The Counter (OTC) Share Market for new technology-based firms, responsible for identifying and assessing young tech companies, assisting in the preparations of those publicly floated on the London OTC, and serving as a Non-Executive Director on their Boards.

Stephen also became the first Technical Director of International Communications Technology Holdings SA, which was listed in Luxembourg. He says: “We raised capital to fund a multi-function telephone, which uniquely had videotex built into it, through Tring Hall Securities, who were a big broker on the London Stock Exchange’s Unlisted Securities Market (USM) at the time. We produced prototypes but never got further than that because we could never get second-stage funding.”

In 1978/9 with Margaret Thatcher in office, Stephen started working on the Electronic Entrepreneurs’ Forum, arranging meetings in Whitehall between the DTI, “technology enthusiasts”, a range of companies that were trying to raise funds, and potential bankers. Stephen says: “Out of that actually came the Loan Guarantee Scheme, and a couple of other schemes, including the Telecoms Product Scheme, which I myself later took advantage of and produced the UK and Europe’s first software-controlled modem. We had a fully-functional manufacturing prototype, which operated separately with all the Personal and Home Computers on the market at the time, by the then novel process of simply loading from digital cassette their individual respective software ‘driver’ programs. However, I spent most of 1984-1985 trying to, but never could, get secondary financing for it, so it folded.  I should have gone over to California much earlier in my career!”

From 1986-1991 Stephen was Business Development Consultant to Government National Accounts of BT plc as client, responsible for conceiving and leading the launch of ACCESS UK, the cross-industry and travel/hospitality community inward-tourism online booking and information service, serving on its Board as Founder Technical Director under its Chairman Sir Colin Marshall (then CEO of British Airways plc).

In 1996, his Digital Experimental Services Amendment was introduced to the Broadcasting Bill at its Committee Stage in Parliament:  “I pay tribute to Dr Stephen Castell, of Channel 5 Digital Television… he is a pioneer of digital technology and acknowledged throughout the industry as being immensely knowledgeable on the subject.  I could not have tabled the new clause without his help”, Rt Hon John Whittingdale OBE MP (Hansard, 13 June 1996).

In 2004, Stephen was a Medallist, IT CONSULTANT OF THE YEAR, in the British Computer Society IT Professional Awards.

For the last three decades, Stephen has been an expert witness in court cases where companies have turned to independent dispute resolution and legal recourse to settle the outcome and ultimate responsibility for failed IT systems and projects. He has established an international reputation and standing in particular as an expert witness in major complex computer software and systems disputes, including the largest and longest such legal actions to have reached the English High Court (AirTours v EDS, 2001; GEC-Marconi v LFCDA, 1992), and Sydney Supreme Court (ITSL & ERG v PTTC, 2012), in IP cases (patent, software copyright: eg USA cases such as BI v Echostar, 2002, and Lodsys v Kaspersky, 2013) and commercial confidentiality actions (eg the Leading Employment Law ‘What is a trade secret?’ Case, FSS Travel & Leisure Systems v Johnson, 1998, the English Court of Appeal), in data forensics and e-document authentication, and in software and technology valuation and quantum cases.  His seminal paper ‘Forensic Systems Analysis: A Methodology for Assessment and Avoidance of IT Disasters and Disputes’ is a Cutter Consortium Executive Report, Enterprise Risk Management & Governance Advisory Service series (Vol. 3, No. 2, March 8, 2006).

From his time as an expert he says that every case is novel. He says: “It’s the same in a different way each time, but there are some very common signatures about why projects fail.

“If you take a bespoke system, and even with ERP systems, and package systems, you still get the same issues…  Because you usually end up tailoring them or customising them.  You end up with this tension between the definition of the requirement and the code written to satisfy that requirement. Then there may be issues about testing and configuring and performing that code, plus the project management of the delivery; because it’s a construction project. Often, cases make allegations more about the way the project was managed rather than necessarily the quality of the code to meet the requirement.”

Stephen explains that quality legally is defined as fitness for purpose, which is the key concept at the heart of many software and systems legal actions.

He goes on to say: “There’s another big area that’s become more common in later developments in the industry where we’ve had much more package code customising wholesale replacement of legacy systems by so-called new techniques.”

He continues: “There’s a lot more today in the whole area of what you might call forensic data analysis to do with cybersecurity, and privacy, and with authentication of documents.  There’s a much bigger mixture of types of case, not just the classic contract for ‘I want a system and it goes wrong because it’s a construction project’.”

He adds: “The great mantra is that you don’t go to court for law specifically; you go to court for money.  And, unless you can show that there is consequence, then what are you doing in court? You may have 1,000 defects, things that don’t work as they were supposed to when they were specified. If none of them work, what would actually be the damage to your business?  For 950 of them, probably very little, but hang on a minute, there are those remaining 50 – and so you get this idea of a material defect.”

Stephen and his team evolved for the first time an objective model for a software material defect. He explains: “It had never been done before; and we showed that there are two orthogonal measures, i.e. key factors quite independent of each other.  One orthogonal measure is the difficulty to fix it because the thing about constructing software, unlike buildings, is that you can ‘rebuild the building’ every night, chuck the old ‘software bricks’ away and start again the next day. The software brick is virtual and it’s infinitely reproducible, so, the time to fix is just a matter of rebuilding the building, as often as you like.  But there could be different levels of difficulty from pretty easy to impossible, because it may be that it’s a design fault; nothing to do with the actual code. Well, one can’t easily fix that, because the thing was never designed that way in the first place; and there’s another whole area of difficulty to fix in regard to: how do you find and reproduce the defect and identify the cause of it? The other orthogonal measure is the consequential effect of the defect: what (financial or other quantified) damage would this defect cause to the business or user organisation if it were to persist in the live operational system?  So, we eventually said that you can only really talk about a software material defect in both the technical and the legal sense if it scores high on both orthogonal measures. If it is both difficult or nearly impossible to fix (and that’s a matter of judgement as to what that actually means for any particular code and any particular project), and is also of major consequential effect, then you can talk about software material defect. In fact, that maps well into the code of a contract and a how you terminate a contract.”

On the perception that there are more public sector IT project failures than in the private sector, Stephen says: “Often in the public sector, it’s not so much that they fail, as that they just terminate.  They just get cancelled because they’re running on too long and spending far too much money.  There has been a common feature that the public sector procurement of IT has not been of the best, to say the least.”

Stephen believes that commercial companies have an advantage over the public sector when it comes to procurement because they better understand the commercial world and generally have their own expertise to scrutinise the sales pitch, project management, and delivery, whereas the public sector does not have the same commercial experience-based insight and guidance. He points to the Central Computer and Telecommunications Agency which used to provide at least some best practice guidelines about procuring IT systems, but which has long since been closed down and replaced by the Office of Government Procurement. He says: “I think this is the problem, there’s no centre of excellence within government.”

Stephen goes on to say that, furthermore, the problems of IT systems, software and data security and reliability, are essentially not fixable. He explains: “I carried out a major study commissioned by the CCTA (H M Treasury) on the admissibility of computer evidence in court and the legal reliability/security of IT systems, still seen by many as a definitive in the field, a sanitized version of which was published as The APPEAL Report (1990, May, Eclipse Publications, ISBN 1-870771-03-6).  This concluded with what became known as: Castell’s (First) Dictum: “You cannot secure an ontologically unreliable technology by use of an ontologically unreliable technology”.  So, as I said right at the start of the whole cyber security industry: there’s no such thing as computer security.  You’re simply building houses out of straw and then putting walls of straw around them.  It might keep people out for a while, but it’s still ‘all straw’. That’s the trouble; the technology to protect computers is exactly the same technology as the thing you are protecting.  …  And of course the only thing you can say with certainty about software is that it’s uncertain.”

Electronic Transactions and Evidence are acknowledged to be based on the concept of a transactional chain of trust.  Stephen identified in 1993 the latter’s dependency on Trusted Third Party Services (‘TTPs’): ‘A Trusted Third Party is an impartial organization delivering business confidence, through commercial and technical security features, to an electronic transaction. …’, noting that TTPs are provided and underwritten not only by technical, but also by legal, financial, and structural means (S. Castell, Code of practice and management guidelines for trusted third party services, INFOSEC Project Report S2101/02, 1993; Commission of the European Community, Green paper on the security of information systems, ver. 4.2.1, 1994).

More recently, Stephen was Expert Contributor to the 2016 Draft Convention on Electronic Evidence, the first ever treaty dealing with the status, examination, recognition, and admissibility of electronic evidence.  The Draft Convention was initiated and developed by Stephen Mason, Barrister and Author, and published as a supplement to the Volume 13: 2016 issue of the Digital Evidence and Electronic Signature Law Review.

Stephen is author of the much-cited article, ‘The future decisions of RoboJudge HHJ Arthur Ian Blockchain: Dread, delight or derision?’, Castell, S. (2018), Computer Law & Security Review, Volume 34, Issue 4, August 2018, Pages 739-753, the Landmark 200th issue of CLSR under the Editorship of Emeritus Professor Steve Saxby.  This noted that, while many are concerned about defining and developing AI ‘Machine Ethics’, Castell’s Second Dictum: “You cannot construct an algorithm that will reliably decide whether or not any algorithm is ethical” (2017) reveals that this is a futile exercise.  “Talking about the ethics of machines might be like speaking of the happiness of water” (page 743).

Stephen’s October 2017 proposal to HM Treasury that the UK Government make a cryptocurrency Initial Coin Offering, for perhaps £2bn, is currently being pursued with the Digital Currencies Team, Bank of England.  His proposal is to issue a new BOECoin: (i) redeemable for real GBP coins and notes, at certain fixed rates initially; (ii) HMRC will accept payment of certain taxes in BOECoin; and (iii) this is non-inflationary, and a direct way of reducing the Money Supply.  He features in appearances on https://cryptoblockchaintv.com/ (for example, in Episode 01).

In his letter ‘What the ECU stands for’, published in Computing magazine 20 July 1995, Stephen proposed that “As cybertrading grows, the new, powerful, common electronic trading currency will be ‘owned’ by no single physical nation state, central bank institution, economic or political grouping. … the ECU … the Electronic Cash Unit”.  This was well before the Satoshi Nakamoto paper ‘Bitcoin: A Peer-to-Peer Electronic Cash System’ of 31 October 2008.

Interviewed by: Richard Sharpe on the 15th November 2019 at the WCIT

Transcribed by: Susan Hutton

Abstracted by: Lynda Feeley