The answers to these questions may come partly from the law

• but only partly!

• There is also a religious or social code
  - and very often also a highly personal code

Ethical judgements are very demanding and the arguments can be very confused!

Sometimes, for example, our views are prejudiced in ways we can scarcely imagine

• Maybe from history?
  • Perhaps even from the Declaration of Arbroath of 1320?
    "It is not for glory or riches or honours we fight, but for liberty alone, which no good man gives up except at length with his life"

• Or from our upbringing?
  • whence do our usually highly emotional views on countryside access or on foxhunting derive?

• Perhaps from our schooling?
  • Are we accustomed to sexual or religious segregations and divisions in society?

• Or even from our friends?
  • Do they alone account for our taste for blue or orange or green?

We need to be very aware of these unconscious prejudices and of the way they affect our judgements -

For evil to flourish, said Edmund Burke the great eighteenth century philosopher and parliamentarian, it is only necessary that good men do nothing.

§7.2 Ethics - A Better Defined Base

Leaving the ethical high ground (or at least the emotional high ground!)

• what do we mean by "ethics"?

In simple terms, the answer to this is easy: doing the right thing

• but, then, who defines "right"?

  Juvenal raised a related and well known question nearly 2000 years ago
  Quis custodiet ipsos custodes? - "Who guards the guards themselves?"

Now Juvenal's question was in one sense flippant: he was making a political point, by arguing that, for preservation of his wife's chasitity, a chastity belt was essential, since otherwise she would be able to seduce her guards. But the cynics tell us that every man has his price; and down the centuries Juvenal's question remains unanswerable. And equally unaswerable is the question of who should define right!

Ethical theory assumes a world in which people

1. are rational

2. make free choices

Beware! Both assumptions are risky

• people can be emotional

• and people can make mistakes

Both emotion and mistakes are common, are indeed very much a characteristic of humanity, and both can thus violate our simple assumption.

• But normally, ethical behaviour is not burdensome:
  - normally, we are honest

  - normally, we do not cheat

  - normally, we do not lie

  - normally, we do not steal

A follower of political fashion may recognise in this the current vogue for vetting administrative systems by way of "a light touch", just as those familiar with Christianity may recognise echoes from Saint John's Gospel, "My yoke is light". In either case the intention of the key phrase is the same, to demonstrate that the activity proposed is wholly ethical, is not burdensome, is - indeed - welcome. And that leads us on, to our next general point:

• If ethical rules are generally recognised as valid, they will normally work
  - ethical behaviour is "good" behaviour

  - and, as such, often rewarded

Note the possibility of devastating exceptions to what is now almost a commentary on behaviour

• ethical behaviour leading to personal loss
  • a mild flavour of this can be found early in Shakespeare's Richard II, when John of Gaunt, having voted to exile his son, cries to the king,
    "You urg'd me as a judge; but I had rather, You would have bid me argue like a father"

• or "general recognition" driven not by sense but by moba emotion
  • where examples are depressingly common, for example the stimulated Kristallnacht of Hitler's brownshirts and others in Germany in 1938, or the "ethnic cleansing" carried out in the 1990's in what was once peaceful and beautiful Jugoslavia

But if the base concepts are reasonableness and acceptability, how do we go forward to build some sort of formal system to allow us to regulate our lives in a rational and ethical fashion?

One approach to ethics is the construction of a set of "good" rules
- for example Immanuel Kant (1724-1804) argued (amongst many other points!)

• that if something is right (or wrong), then it is always right (or wrong)
  Kant's Principle of Universality

• that if something is logical then it is right
  in Kant's words, "Respect the reason in you"
  meaning, of course, that we should use knowledge and judgement and not raw emotion

• that people are [always] ends in themselves, never means to an end

Alas, even the first of these has a flaw; just ponder the possible reactions to the statement, "Please, tell me where is the person I wish to kill."

Another approach, and one very different from Kant's rigid framework of rules, is the construction of a regime of consequences

• of which John Stuart Mill (1806-1873) gives us the best known example, in his development of "utilitarianism"

• it was Mill's belief that we should work to increase people's happiness,
  choosing decisions which maximise the sum total of happiness

• for anyone who has had occasion to study thermodynamics, a theory developed at the same general time, there is a nice parallel with the thermodynamicist's views on entropy!

Alas, there is no simple answer or frame work

• Socrates (469BC-399) is reckoned the first ethicist ...

  ... and he was sentenced to death for corrupting the young!

So, what do we have for a basis for our ethical considerations? The fundamental points are very simple, and yet very hard to define; we have:

And we can conclude (remember the earlier quotation from Edmund Burke?) that we have responsibilities

• to provide certain things for others

• to act independently, free from pressure

We note the difference between "wrong" and consequential harm

• creating excellent software will cause your competitors harm as they go broke ...

  ... but that creation is hardly unethical!

• conversely, hackers who do no harm ...

  ... are wrong in ethical terms (and also, a very different matter, at law)

A goal is not necessarily bad of itself, but the context applied to get there can make it so; thus

• maximise profit
  • by hard work ... fine!

  • by lying and stealing ... very seldom acceptable!

Laws and ethics are two very different systems, though there may be major overlap between their ideas, for the simple reason that codes of law usually reflect the opinions of society, whilst codes of ethics normally are those opinions.

§7.3 Applications to Computer Systems

If we try to apply these thoughts on ethics to computing systems, what do we get? What forms a topic range for "computing ethics"?

Probably we need to restrict the scope of the term, to narrow it down from points of social or political opinion

- for example, to rule out the issue of working with computers in the defence industry
- working in the defence industry, or not, is (to the class lecturer's mind) a valid ethical worry
... but it is not (again in the lecturer's view) a matter of computer ethics

But what we can more easily do, is to draw out some guidelines likely to leave us, as computer scientists, with ... well, with ethically clean hands.

Thus we might arrive at such issues as:

• Fully understanding project goals
  • not a working computer system, rather the project of which our computing is a part
  • for a contrary example, find out about the London ambulance fiasco of the mid-1990's

• Including real users in test regime
  • which allows us, for example, to verify test data values and sources, or establish relevant accuracy levels
  • for example, the premature baby problem ... a beautiful computer system, for helping diagnose the problems of such tiny babies,
    but it only recorded the weights of those tiny babies in integer pounds, not in the small fractions - ounces - which were required

• Planning and scheduling
  • need for accuracy

• Designing for real (and therefore imperfect!) users
  • "to err is to be a user"
  • your system must handle those user errors

• Never assuming that existing systems are necessarily either safe or correct

• Always requiring a positive case for system safety
  • "seems to do no harm", is wholly insufficient!

• Being honest about your system's limitations and shortcomings
  • to do otherwise, is simply evidence of poor testing!

All these points can be developed from one or other of BCS' codes of practice and of conduct - or from equivalent codes published elsewhere, for example in America by ACM or by IEEE CS.

Note that the ethical problems which emerge tend to blend with technical problems or (though not in the examples quoted) with legal ones - the area to which we will be turning in the next chapter - but it is important, and down to us as professionals, to keep in view not only the technical and legal issues but also the ethical ones.

§7.4 Footnotes

Four points before we move on ...

The first concerns research which involves anything affecting humans

• In the University, there is an "Ethics Committee"

• And, associated with it, there is a required departmental structure ...
  ... with responsibility for approving all departmental work involving human subjects
  • "all work" is not restricted to the physically intrusive, it extends to psychological factors

  • and thus to the professionally required evaluation of any form of project ...

  • ... don't forget when you get to your final year project!

• It is now becoming clear that the UK government will be enforcing a similar pattern on all research work
  • But even some years after the introduction of the scheme (late 2003) the significance and implications of this remain far from clear.

The second point? Ah, a point for all you who thought you could sidestep these issues because you were going to work in management rather than in the more technical areas of computer science. That doesn't let you off the hook:

• A TBS graduate of not so long ago (graduated summer 2003) has pointed out how, before he was allowed to take up his employment with one of the country's leading financial firms, he had to read and sign that he would abide by an eighty page ethical code.

Thirdly, if some of these ideas have struck you as matters of curiosity in their own right, you might enjoy the very much wider study (wider than simply ethics, that is) to be found in the following book:

Though do be warned: it is an excellent and fascinating book, but it is also a wholly American book! You may also be interested in - still American, but less stridently so -

which may well form a - or even the - recommended text book for future versions of this class.

The fourth point is rather different. This chapter has been unusually free with the sorts of quotations not often found in Computer Science lectures. So let us end with two more quotations, and with a moral. The essence of ethics is the willingness of each individual to think, and to be willing to think for him or herself. It is that individualism which is the ultimate source of political freedom, something which even in the twenty-first century a large part of the world population does not enjoy. The erstwhile British leader Winston Churchill put one side of that necessity to think:

And where are Churchill's words and thoughts traditionally stored? In books. And throughout the ages, to ensure conformity of thought, there have been book burnings, not least those by the Nazis in May 1933 in Berlin's once again beautiful Unter den Linden. Over a century before, speaking of a burning of the Qur'an during the sixteenth century Spanish Inquisition, the German poet Heine had said

That quotation has been heard at many recent Strathclyde graduation ceremonies: it is our responsibility to treasure knowledge, and to continue to work at our own freedom to think.