Gilbert Ryle famously uses the institution of a university to illustrate a category error, in which a university is confused with its visible parts. The university is the word, according to Ryle, that describes the organisation of those parts, and to see it as a thing in the same way as its parts are things is wrong.
A foreigner visiting Oxford or Cambridge for the first time is shown a number of colleges, libraries, playing fields, museums, scientific departments and administrative offices. He then asks “But where is the University? I have seen where the members of the Colleges live, where the Registrar works, where the scientists experiment and the rest. But I have not yet seen the University in which reside and work the members of your University.” It has then to be explained to him that the University is not another collateral institution, some ulterior counterpart to the colleges, laboratories and offices which he has seen. The University is just the way in which all that he has already seen is organized. When they are seen and when their coordination is understood, the University has been seen.
(Ryle 1949, 17–18)
But Ryle’s university problem also touches on the specific ontological weirdness of any institution. What is an institution? It might be the sum of some parts – but what are those constituent parts? And if it is specifically the organisation of those parts, what is the nature of that organisation?
Is the university constituted from colleges and libraries, or from the people teaching and learning inside them? Ryle gave us some locations that would not have existed when Oxford or Cambridge universities first came into being. Oxford’s claim to being the oldest university in the English-speaking world is not based on the date of construction of its famous buildings, which turned up centuries later, but on records of teaching happening in the late 11th century. The university came about because a bunch of scholars got together to teach and advance knowledge; it continues to exist for that purpose, albeit in a larger form, and it can be physically located in a swathe of material objects – all those famous buildings – that it built for that purpose over time.
I have no idea about the intricacies of how universities are run, or what sort of written constitutions they have, but will continue with Ryle’s example and tweak it in order to carry out some thought-experiments. The fictional University mentioned henceforth is a venerable institution founded, like Oxford, by some monks who intended to disseminate and advance knowledge, and at some point down the line that purpose was set out by a particularly literate monk into a written constitution.
Let’s consider two possible events that might take place. In the first, a zeal for privatisation sweeps the British university sector. The old, rich, well-regarded University is in a strong position to detach itself from any funding from the state and sets up as a private university. Its purpose remains to disseminate ideas, but it can do that and generate more money in order to grow as an institution. It is still the University: it does all the things a university does, employs its fellows and professors and administrative staff, and carries on educating people and advancing knowledge. It just has more money to throw at it now.
At some point down the line, it gets taken over by EliteUniCorp, a vast multinational education company promising to inject additional capital and resources into disseminating and advancing knowledge. There is a handover period of an academic year during which EliteUniCorp legally owns the University but makes no managerial changes, preferring to observe it first to work out how to make some money from it. The University looks like the University still – it has retained its valuable branding despite being subsumed into being part of EliteUniCorp, whose purpose is to make money for its shareholders by selling education and research services. Ryle describes the institution of the University as ‘the way in which what [the visitor] has seen is organised’: a visitor wandering past at this point would see the activities characteristic of a university, because in those famous buildings, students are being taught, and books are being read, and monographs written, and advances in knowledge continue to take place. That activity is what is being organised. EliteUniCorp see it as a nascent business. Is it still the University? Some left-leaning professors do not think so, and express concern on Twitter, citing abandonment of the aims of the University’s constitution in favour of making money for EliteUniCorp as evidence of its demise with the hashtag #deathoftheUniversity. EliteUniCorp responds with a statement that nothing relating to the university activities of advancing and disseminating knowledge has changed, and begins to transfer the University’s profits to its own accounts.
In the years that follow, EliteUniCorp gets greedy, and cuts professors’ salaries while charging ever more outlandish tuition fees that narrow the field of people willing to pay to a handful of oligarch kids who are lazy and not very bright. The professors leave in droves, the reputation of the University suffers and soon EliteUniCorp struggles to recruit both staff and students. Once it finds the University running at a loss, EliteUniCorp shuts down its University operation. No point in educating people and advancing knowledge if it doesn’t pay. The buildings are still there, and busloads of tourists still come to the city to see them and take photos of them, and can even pay to stay in them now that they have been recreated as Brideshead Revisited theme hotels by a Chinese tourism conglomerate, but no teaching or learning is happening any longer.
In a second future world, the Computer Science department proposes to create a digital learning platform that can share the University’s research and teaching with neural networks for mutual benefit. This falls into line with the University’s purpose of teaching and advancing knowledge and the platform goes online; in fact, the neural networks are much, much cleverer than human undergraduates, and soon enough colleges compete to sign them up as student-entities in order to improve their ranking in intercollegiate results tables. Soon, the only human students left are beefcake brought in for rowing teams, and when the Other University’s new genetically modified lab-grown Row-Bots beat the Blues, they no longer have a function either. The University employs technicians to maintain the servers in the Computer Science department, where the Computer Science department executive neural network communicates their daily tasks through an idiot-proof productivity app in order to minimise human error. There are a few human lab technicians left in the departments where robotics have not yet been adequately finessed to take over. The Vice Chancellor neural network has been in place for decades, ever since it was trialled against the erstwhile human Vice-Chancellor and found to be more effective; when the University’s last Chancellor stood down, it was determined that there was no further need for any human to take that ceremonial position, and that ceremonial positions, gowns and fancy dinners no longer have much purpose either. The existence of competing yet collaborating college-nodes turned out to be a highly functional networking structure, so the various colleges still exist, and still compete to produce and disseminate the most information, but the buildings got sold off to that Chinese tourism conglomerate last year and are now hotels. When those tourists go to the Bodleian Wellness Spa or Brideshead Experience, do they pay all that money because they commit the marketing-assisted category error of thinking that they are visiting the University?
An institution isn’t, in these examples, the same thing as its constituent buildings. Is it the same thing as its people? It doesn’t have to be: the AI university seems to be doing perfectly well at the core activities of a university – disseminating and advancing knowledge – activities that those eleventh-century monks wished to get together to promote. The monks might not have considered it possible for the university to educate, or be run by, non-monks; it might be no less strange or heretical to them that its activities be carried out by computational devices and robots than by women and non-Christians.
In that case, where do we draw the line regarding where the institution of the University exists, and where it doesn’t? Does it still exist when it is teaching people and advancing knowledge, but to make a profit rather than for those purposes themselves?
When the University first goes private in the example above, its core purpose remains to disseminate and advance knowledge; the financial systems it employs to do so are not driving its purpose, but supporting it. Once it gets bought up by EliteUniCorp, I’m with the hypothetical dissident professors: it no longer exists as an institution. It becomes part of another institution instead. Its original purpose has been supplanted by the purpose of making a profit by supplying teaching and knowledge advancement services, and while the machinery of teaching and learning is still in place, it no longer operates according to the principles of the Oxford University constitution – to disseminate and advance knowledge – but according to the principles of EliteUniCorp. An institution is an entity made up of other entities organised to carry out a common aim, and once the machinery of the University is carrying out the aims of EliteUniCorp instead, it becomes the machinery of EliteUniCorp. The University is no longer a university, but an arm of an education services company.
In the time when the University goes private before getting bought up, it might have to amend its constitution to allow for its new financial systems. I would argue that it is still the same institution, because it retains its constitution and its aims, but tweaks them slightly. When it decides to move towards disseminating knowledge via AI rather than human brains, it might need to amend its constitution in case there is any pesky out-of-date anthropocentric bigotry holding it back from admitting the intelligent entities purely on merit, just as it once amended its rules to admit women.
The institution is, then, the collection of entities that exist to carry out a common constitution. The constitution does not necessarily have to be formally written down: I might start a book group whose aim is to meet on Tuesday nights in a pub to talk about a book we’ve read in common. At some point, I cannot get a babysitter and stop going, but the others still go, and books still get read and discussed in the pub, and a year later it’s going strong without any of its initial members. It’s an institution, says the pub landlady, grateful for the custom on a quiet night. It might be agreed one Tuesday night that it is, in fact, better to meet on Thursday nights instead, and it continues to exist with that alteration. It could move to another pub a few months later after falling out with the landlady, and continue to exist. She doesn’t own it, and nor does the pub, and nor do I. It is more than its physical location, and the people who turn up to read the books, and the initial rules for when and where it happens. It exists for as long as there are people who agree to carry out its aims, and even when some of the rules for how it happens change, its existence is sustained.
The book group doesn’t have a written constitution. It’s more a loose agreement, a mutual understanding. People who want to read a common book and talk about it in a particular place and at a particular time can come along and do so, and there they will arrange the next meeting. The book group, as an institution, might be the way those common activities are organised into a single event: book, discussion, pub, Tuesday. A regular observer in the corner of the pub would see that pattern from week to week and conclude, if he was not too drunk, that people turning up to talk about a book was not merely a random crossing-paths of bookish folk who happened to have the same book with them, but something organised. If another punter entered the pub and asked the landlady to show him the book group, she would point towards the people in the book group, but it would be hard to make the category error that the university-seeker made: it is visibly a group with books. And yet the book group is not dependent on any of those individual people for its existence, because all the original members have left.
If we are going to pin down what defines an institution, the constitution looks like the best bet. Ryle’s description of the university is ‘the way in which [its constituent parts] have been organised’ could be taken to mean a constitution. There might be formal records of a constitution, as large organisations like universities and nation states tend to have, or there might not, as with the book group. If, as the founder of the book group, I were to insist on agreeing a formal written constitution with the other book group members, they would probably run for the door.
We could look at the constitution in two ways. In the how version, it might be described Ryle-style as ‘the way in which an institution’s constituent parts are organised.’ The how constitution would describe the layout of the university’s knowledge economy, employing professors to acquire new knowledge and disseminate that knowledge to each other and to students. It would provide job descriptions to the people carrying out activities within it, and set out meetings between colleges and faculties. The how of the book group would be a common procedural understanding: its members will read the book chosen by majority agreement at the end of this week and turn up next week in the pub to talk about it.
The why constitution would look more teleological: it would be a set of shared aims that, in turn, drive its various activities. The University exists to advance and disseminate knowledge, and each department and individual within it acts in order to support those aims. I am highly sceptical of mission statements and the people who write them, and indeed the term ‘mission statement’, but it would basically be just that. The Book Group exists to facilitate discussion of books between friends in the pub.
Both of those constitutional descriptions are, in my opinion, useful and interesting. Either way, the point at which the institution ceases to exist must be when it stops carrying out its constitution. Whether or not we take a how or why version of its constitution might align with where we stand on the dissident professors’ claim — that the University ceases to be the University and becomes part of EliteUniCorp as soon as it gets bought*. In the how version, the University would cease to be the University at the point where it undergoes observable procedural changes: you might argue that this happens as soon as its financial systems are updated to send profits to EliteUniCorp, even if at that point the how of the day-to-day organisation of university activities would be identical to any observer. Perhaps it would happen later, when the admissions guidelines are adjusted to accommodate the growing numbers of the rich-but-thick, and a new Student Behaviour Alignment department is put in place, which collaborates with the new Fast Track Gold office to encourage students to pay extra for better grades rather than intimidate their professors.
(*I was going to say that it depends on where we stand on the dissident professors’ claim, but then I realised that I think they are right because I take a why position, rather than taking a why position because I think they are right, and will explain this in the next part of this essay.)
The why version is more clear-cut: once the University no longer carries out its constitutional aims, it stops existing. There, the constitution is the set of aims that drive institutional activity. The activity might look the same, but if it is no longer driven by the aims agreed upon within the constitution, some other entity is doing the activity – in this case, EliteUniCorp.
The why version need not negate the how version, and vice versa. A key difference is that, in the why version, the constitution is more than an observable pattern of organisation constraining and shaping institutional activity. It is a sort of entity in its own right – the category error, so far as Ryle is concerned, of holding the University as a thing. But it doesn’t have to be a formal, written entity; it can be an agreement, or an alignment of aims.
If the why version of the constitution is made of anything, it is made of ideas. And what is the ontological status of those? I can’t see how you can keep going without recourse to either dualism or idealism in some form.
As for the how version, you could argue that there is nothing woo-woo going on and that the constitution is merely an organisational pattern, but the nature of organisational patterns is that they constrain activities so that, for example, people at the book group talk about books rather than watch the football, and people in universities carry out activities relating to knowledge rather than fighting wars, even though it may sometimes seem as though anything is possible. The Army is an institution whose legitimate activity does include fighting wars. In a formal institution with formal rules like the Army, or, less rigidly, the University, individual members would be disciplined if found carrying out activities that contravene its constitution. If a platoon of soldiers were to be found discussing poetry on the battlefield, they would face a disciplinary tribunal, and so would a student found designing bombs in the Biochemistry department. The constraints are not limited to people, or even putative artificial intelligence networks: if the University finance department does not acquire money in grants and fees, and send money to other departments to buy equipment, maintain buildings and pay staff, the University will not last long.
If an institution is the organisation of a set of other entities that carries out some activity, and it can’t organise activity without constraints, what exactly are those constraints?
In the case of the Army, it is easy enough to imagine that the presence of commanding officers with rifles would make you stay on task. Perhaps the presence of men with guns can be predictably seen to impose behavioural patterns on people, especially those who have undergone behavioural training to take orders and carry out particular tasks in a particular way. But those constraints are mostly communicated as rules – in writing, or verbally, or as an observed social pattern. If someone in the book group talks for ten minutes about the football match at the other end of the bar, the frosty reception he gets would probably communicate that he should talk about football somewhere else, and if he kept going for twenty minutes someone would probably pipe up and say that everyone else is there to talk about books. Those rules exist, formally or informally.
What are those rules? You could say the rules aren’t a thing, and to see them as such is a category error, and argue that they simply describe the operational process, or the pattern of organisation. But in most institutions, deviations from operational process both happen and get corrected. Perhaps the word ‘rule’ is a folk-psychological delusion to describe a hardwired group dynamic, but it starts to feel as though we are doing a lot of work to explain away the existence of rules, and the only reason I can think of to do so is that you then face the same ontological problem of what, exactly, rules are. If the rule contravened by the soldiers discussing poetry in the battlefield is ‘in the battlefield, we fight’ or some variation thereof – it might be ‘obey orders’ but the order to their commanding officer is to fight, or it might be that he too is merely obeying orders and the order given by the government at war is for the army to fight – we might reasonably infer that the rule is to fight in that situation, as might a soldier court-martialled a few times for not fighting in a battlefield – if he managed to stay in the army and/or alive for that long.
How different is a rule from an aim, abstraction aside? And then we are back in a similar domain of aims, ideas, etc as we were with the why constitution. You could avoid arguing that rules exist to constrain behaviour to a particular end, and simply state that they constrain behaviour in a particular way, but those constraints still have a specifically informational character that needs to be disseminated among the constituent parts of an institution in order to ensure that the activities it does take place. And I just cannot work out how you get rid of the problem of that information, or idea, or whatever you want to call it. Perhaps you could describe it as a process or a pattern and try not to situate it as an entity aside from the entities in which it operates, but I just used the word ‘operates’ as a way of trying to get around the existence of a noun, and the process or pattern or whatever it is has a particular effect, or, if that sounds too teleological, makes a particular shape. The language-play seems pointless: each time, we end up saying the same thing in a more circumlocutory way: why not just accept that information exists and ideas exist and they make stuff happen? It’s a whole lot neater that way.
If Ryle was so keen to kill the ghost in the machine, that particular example strikes me as a risky way of doing it. Institutions are weird. The things that make them institutions look worryingly immaterial. If you tell people to stop erroneously thinking of them as entities and rename them organisations of other entities instead, the haunting doesn’t magically disappear. An organisation is, after all, organised.