4 days in Edinburgh

image

(Not shown: It Might Get Ugly late on the Saturday night)

Sh*t-faced Shakespeare

This is a show with a very simple premise: take a cast of 6 people, condense a Shakespeare play down to an hour, prepare it as a serious Shakespeare production, then pick one of the cast and get them very very drunk before the play starts. Then you have 5 people trying very hard to do the play properly and one person who is completely incompetent: coming onto stage before their cues, missing the cues, telling other people to kiss, telling the audience that she’s going to do some acting now, and generally messing stuff up. It was good fun but not super-hilarious. How good it is probably depends a lot on who they get drunk and what whims flit across that person’s mind on the night.

Boris & Sergey’s Astonishing Freakatorium

Really impressive puppetry: most of the puppets took three people to operate (one person on legs, one on arms, one on head, I think) and they were able to use that to give the puppets a really expressive physicality. It also meant that despite having 6 puppeteers they were limited to two characters on stage at a time, which is a bit ridiculous. They did have a fun (if very ambitious) audience participation bit where they called upon two volunteers to assume the characters of spirits summoned in a séance. The puppetry was technically amazaing and I was charmed by the two main characters, but on the whole the show was a bit excessively bleak.

Abandoman: Hot Desk

Good fun, but not the heights of brilliance I was hoping for (this show cane highly recommended by several people). The setup is that the cast have to write a hit song in the next hour or one of them will be killed. We the audience supply ideas that they then improvise a rap around. They start with some love songs (one about a couple in the audience who they quizzed, one about someone’s crush on Lara Croft), then some songs for old people (about hobbies, investments AMD reminiscences), then a song about a first world problem (soggy croutons, in our case). The guy worked the crowd incredibly well, but the show didn’t really click for me.

The Horne Section: Milk the Tenderness

I liked this. Alex Horne is an impresario leading a 5 piece band for a sort of variety hour. They have a guest come in and do a bit (we got a standup who ended his bit by remarking to Horne: “you know, I don’t think I’ve made an audience that angry with for a very long time”. The bit was a shambolic one but I couldn’t work out if it was meant to be or not). Anyway, Horne and the band do a bunch of musical comedy bits and Horne drives the whole thing along brilliantly.

Viewmaster

This was good, too. You start in the Summerhall courtyard, which is buzzing with food, drink and merriment. Then you are led down some passageways, past queues of people chatting quietly as they mill around waiting for their show, and up the stairs to the quietness of the loft, perched above and at a remove from all that chaos. You fill in a questionnaire and enter Dan and Ryan’s den. They look at the questionnaire, analyse the results, and pick a journey for you to go on.

Dan plays the music for the journey, Ryan reads the poetry, and a viewmaster provides the visuals of the journey. It’s a beautiful little experience (the whole show is for one person and only takes 15 minutes) that draws you out of Edinburgh and into the locale of the particular journey with remarkable vividness.

Nofit State: Bianco

This was a very intriguing show. I think it was probably the biggest show that I saw, in terms of both ambition and size of cast. It was held in the Nofit State big top, which was operating as a venue without seats. They had four wheeled truss structures that they moved to provide various settings for the action to take place on, and as the action moved the audience was moved to follow it. This was both a strength and a weakness of the show, with quite a lot of the total running time being taken up with rigging changes as they moved stuff around. They were going for this very dynamic thing of the action shifting and changing and the audience coming after it, which was a beautiful dynamic, but it was hindered by the time taken to move stuff around and lock it into place. The overall vibe of the show was wonderful, feeling like a celebration brim-full of joyful chaos (I know I said something similar about Akoreacro. Once is a coincidence. Twice is a genre). And they had some immensely skilled people, particularly a tightrope walker who did a frontsault.

The Jest

The Jest are a five person sketch comedy crew. My relationship with sketch comedy is a complicated one (I don’t usually like it but I keep going to see it). A lot of their stuff was pretty good. I particularly liked their wordplay where they’d pick a particular actor and work a ton of titles of their films into a sentence.

Jamie Adkins: Circus Incognitus

I think the last one-man circus show I saw was the brilliant ‘Kaput’ by Tom Flanagan. This was a reminder that a show that good is quite unusual. There was nothing wrong with this show, but it felt very formulaic and workmanlike.

Mark Grist & MC Mixy: Dead Poets Death Match

This was fun, although very silly. The show ends with the two guys adopting the characters of two dead poets for a rap battle. On the way, the guys talked about various dead poets and the influence they had felt from them, and did poems inspired by those figures. This was kind of in the intersection of spoken word, poetry and rap, and I think I would have liked it to be more on the straight poetry side, but it was a fun and engaging hour based on dead poets.

Torsten the Bareback Saint

This was pretty bad. It was billed as a song cycle about a Dorian Gray figure, which I guess it might have been, but that didn’t come through for me in the songs. For me the songs were a bunch of disparate points that I was unable to link back into that supposed central theme. If the songs had grabbed me, that might not have mattered, but they didn’t. They’d made a weird choice to have a pianist on stage but a lot of the music was coming through on a backing track. Either have a backing track, or have an accompanist, or make a joke out of having both. This show failed there.

And I haven’t even mentioned the biggest problem with the show, which is that it felt like the performer was taking the audience for granted. This was an side project for him, and I suspect he’s used to playing much bigger stages. His mode and manner might work for playing a bigger room but just didn’t translate to the smallish room he was in.

Or maybe he’s just not very good.

Shaun Usher: Letters Live

This was the only non-fringe event I went to, being part of the book festival. It was a curiously lifeless event hosted by two people, one of whom is behind the ‘Letters of Note’ project and the other is the author of a history of letters and letter-writing called ‘To the Letter’. Through their work these guys have discovered a lot of letters written by and to various people, some with great gravitas and some entirely frivolous. The event consisted of a cast of actors, comedians and so on coming up and reading a selection of these letters. It didn’t work for me, with the length of the letters being part of the problem. Much of the beauty of a letter is the intimacy of it, the strength of authorial voice. And here we had people standing up and reading letters that were not their own. The very sterile context of a book festival stage and the rather short length of the letters did not give the readers the time or the space to claim the author’s voice as their own. That the readers were introduced before they read, thus bringing them into focus as themselves when they needed to be present only as the author’s avatar, added to the problem.

Nice idea, though.

It Might Get Ugly

This was an honesty themed stand-up show with a lineup that changes every night. The brief to the comedians is super-simple: be honest to the point of regret. I think it could be really magical when a skilled comedian really crosses the line into over-sharing, because most of them have practiced walking that line. The night I went was pretty good, but nobody really crossed the line. One of them told a lovely story about the lengths he was willing to go to to visit his girlfriend when she goes to study in Oman, one I have forgotten all details except the two words ‘Canadian’ and ‘lovely’, (oh, a third word: ‘bisexual’), and one told a hilarious story about going back to a girl’s house after a gig, going into her room (in her parents house) and seeing posters of himself on the wall, insisting on seeing ID, then sleeping with her. That last guy was the closest to fulfilling the ‘honest to the point of regret’ brief, but still wasn’t really sharing more about his life than I expected him to be. Still, a good night, even if the MC was a pretty awkward dude. He kept talking about how much the producers had paid him to take the show, which is a regular event in London, to Edinburgh. I assume the producers plan on some shows going well and some going poorly, and seeing the crowd this one drew on a Friday night I assume this one falls onto the ‘poor’ bucket. Which is a shame, because as shows with a lineup of comedians go, I’ve seen a lot worse.

And that was my 4 days in Edinburgh.

Advertisements

Progress report

I remain captivated by the idea of a large-scale software system implemented entirely in purely functional style, and have been noodling away for a while on a much simpler implementation of some of this folly.

The old way was to take pure functions and try to magically incrementalize them. The problem with this – apart from it being really quite difficult – was that I don’t want purely functional style all the way through for its own sake. I want to reduce the number of sources of surprise in day-to-day software development. In practice, a magical incrementalizer can only magically incrementalize some functions, and the rules around which functions get nice big-O behaviour in the incrementalized versions are surprising indeed. So maybe that’s not such a good idea.

Instead of writing pure code and turning that into incremental code, I’m now writing code in terms of a set of primitives (lists, maps, structs) that know about acting incrementally. It works pretty well and I’ve just finished a big reworking so the guts of the thing are much simpler.

The implementation takes the user’s function (written in terms of the below primitives), and turns it into a network. The difference between this and every other FRP or related system I know of is that the change propagation network is completely distinct from the value network, and the primitives and operations are designed so that any change in the input can be pushed through to the system state with nice big-O behaviour. This is accomplished by evaluating the user’s function once – with a funny notion of evaluation – then walking all changes through the resulting network.

Every invocation of a primitive becomes a node in the network. Each node has some inner nodes; changes are propagated from each node to the nodes that have it as an inner node, and a node is only fully compiled once its inner nodes have been compiled. In the current design, change propagation goes like this:

  1. A change arrives at a node
  2. The node takes the change and produces a (possibly-empty) list of changes that result
  3. The new list of changes are propagated from the node.

This is likely to change in a future reworking so that a node can propagate a change differently according to where it arrived from (in the current design, StructElems exist solely to work around this lack).

A change is an impulse applied at a location. A location is a path through the structs and maps that make up the value being changed. Once a change is propagated up to the top-level value, it is ready to be applied.

Primitives and Operations

Currently the following primitives and operations exist:

Lists

With lists, the overall list is a changeable evolving thing, but the individual values are not.

map :: (a -> b) -> [a] -> [b]

If the change is adding an element e, map produces a change that adds the element created by applying e to the function. If the change is removing an element e, map produces a change that removes the element created by applying e to the function.

filter :: (a -> bool) -> [a] -> [a]

For both adding and removing an element e, when applying e to the function returns true filter returns the same change. It produces no changes (effectively swallowing its input) if applying e to the function returns false.

shuffle :: (a -> b) -> [a] -> Map b [a]

Shuffle takes a list and sorts it into buckets. For both adding and removing an element e, shuffle causes that change to happen at a path determined by passing e to the function.

inputList

An inputList is necessary to feed the monster. The idea is that you have one (or more, I guess?) streams of inputs to your application. Constructing impulses that add elements to your input list allows you to have your application’s state evolve over time.

Structs

Structs are intended to hold related values. They are largely a way of translating product types into this space of incrementally-evaluated functions. Their structure and keys are fixed, but their values are changeable evolving things.

Dicts

With a dict, the overall structure is a changeable evolving thing, as are the individual values, but the keys are fixed.

map :: (a -> b) -> Dict k a -> Dict k b

This takes a map and applies the function to all the values in the map. I use this to take a shuffled input list and produce a map of structs of interesting things resulting from that list of values.

Interesting aspects of the implementation

There are two wrinkles that might be interesting if you’re trying to do something like this. The first is not a big deal: if you don’t want to split structs into a struct node plus a node for each element, then you need to be able to handle incoming changes differently depending on which node they came from.

The more interesting one is around mapDict. The arguments passed to mapDict are a function f and a dictionary d. The hopefully unsurprising goal is for every elem (k, v) in d to become (k, f v). One approach might be to generate a change propagation network for f for every element in d, but that’s going to get us a huge change propagation network over time. What can we do? It turns out that changes coming out of d are already changes at a location in d. That means that if we construct a single change propagation network for f and feed changes coming out of d into it, we’ll get suitable changes for the mapDict coming out of it.

The question then is how to do that. We can’t just apply dict to f because f is a function from a -> b, and dict is of type dict k a. Therefore, we introduce the ability for the internals to create an arbitrary node of any type, and we create such a node (which we call a) and pass it to f. At this point we have two choices:

  1. Manually introduce the plumbing so that changes coming out of d get propagated to the a, and cause changes coming out of f a to get propagated out of the mapDict call. This turns out to be pretty hairy, but the hairiness is all in MapDict.
  2. Declare that any arbitrary node a has an inner node (which is any watchable thing), and changes in that watchable thing appear as changes in a. This turns out to be a whole lot simpler, with the caveat that each type now needs to not just provide a mechanism for producing an arbitrary node, but also to allow that arbitrary node to have an inner node and compile it appropriately.

Spoiler alert: the second approach is better.

Future Work

There’s so much future work! So much. So much.

Right now, the distinctions between what are changeable evolving things and what aren’t is rather ad-hoc. This could stand some more rigor.

It could be quite useful to have operations like

  • keys :: Map k a -> [k]
  • values :: Map k a -> [a]
  • elems :: Map k a -> [(k, a)]

Intriguingly, the elements of the list returned by values could be changeable evolving things (as could the second elements of elems‘ tuples), unlike other lists. This feeds back into the rigor around distinctions mentioned above.

It would probably be useful to have a way of looking up elements in a struct or map.

The syntax for invoking all of this stuff is hairy. I need either a new language or a much prettier embedding into haskell.

I have integers, but the only operation is addition, and the only way to get one is by summing a list. Generalizing that sum to a fold might be useful.

There’s a growing diamond problem under the hood. I’d like to get that under control, but it feels like it needs to get worse before it can get better. It may not turn out to get worse. There’s definitely some work to be done around removing duplication, especially around inners, ids, and the ability to be a tube.

And all of that is without looking at pushing back towards integration with a datastore or pushing forwards into the browser.

An Oversight

Object Relational Mappers have a dream. Their dream is that the mechanics of creating, reading, updating and deleting can be abstracted away and need not be thought about.

I have a dream, too. My dream is that the fact that creating, inserting, updating and deleting is going on can be abstracted away and need not be thought about.

The ORM dream has a big ol’ flaw, to the extent that I am sceptical about the use of ORMs in general. The flaw is that the right way to do CRUD operations depends rather heavily on the domain and the application in question.

My dream should not be vulnerable to that particular flaw, though, because I’m trying to abstract away those operations entirely. Or so I thought right up until yesterday, when I realised that since my approach is to first convert the programmer’s code into a form that does have CRUD operations, I’m going to have the ORM problem from then onwards.

And it’s worse, because when you call the CRUD operations yourself you can exercise a fair degree of control over what goes on. When the CRUD invocations are automatically generated, that control would have to be exercised by hints and indirect suggestions.

Bugger.

3 Approaches

In today’s fantabulous outing, marvel at three approaches to efficient implementation!  Ride the wave of perplexity that is the function from a list of inputs to the current state! Gaze in wonder at the persistent storage of that current state!

Option 1 – incrementalisation

In this approach, we take a state function f, that takes a list of inputs and produces current state, and produce an incrementalised state function, that takes a change in the list of inputs to the resulting change in the current state. Liu gives a systematic approach for deriving the incrementalised versions of functions.

Advantages

  • It enables a neat way of making a highly responsive UI. As well as incrementalising the state function, you also incrementalise the view function (a function from state to HTML). On the client, you can run the incrementalised state function, pass its output to the incrementalised view function, and apply the resulting change to the DOM. (this works)
  • It lends itself to a really cute way of integrating with a data store: write a function that takes the system state and returns a relational representation of that state. Then incrementalise that function. Any time the system state changes, you can pass that change into the incrementalised relational representation function and hit the database with its output. Shiny! (I haven’t done this yet)

Disadvantages

  • The wording of Liu’s paper is very careful. What they have is a systematic approach for deriving incrementalised versions of functions. What they don’t have is a fully-automated approach. There needs to be some user interaction or some hinting (in the general case) in order to produce an incrementalised version of a function.
  • When you have an inner function that uses free variables, and the values of those free variables can be affected by a change, the incrementalised versions start to get pretty hairy and potentially rather inefficient. The current version of my code assumes that any free variables used will not change, which is obviously not true in general.
  • At the moment I have an informally specified bug ridden implementation of less than half of Liu’s technique.

Option 2 – Partial Evaluation

Partial evaluation is where you take a program of n inputs, then provide some subset of those inputs and evaluate the program with respect to those inputs. This isn’t immediately applicable to this state-updating shenanigans, because there’s no provision there for changing inputs.

Instead of running the partial evaluator directly on a program that takes a list of inputs and returns a state, we can augment the program with a new top level function:

return state(cons(new_input, inputs))

Assuming the partial evaluator unfolds that cons, we’re in with a chance of getting a program that incorporates a new input without doing too much stuffing around. It’s not at all clear how you might get the state into a persistent data store with this approach, but worse then that, the resulting program won’t know how to capitalize on existing bits of state. This is probably surmountable by memoizing computation of bits of the state, but if you’re going to go there…

Option 3 – Adaptive Computation

Adaptive computation (after Acar) is where you think of a program as building up a dependency graph of values. The final output depends on a bunch of intermediate values, which depend on other intermediate values, which ultimately depend on the program’s inputs.

There’s no obvious barrier to storing that whole dependency graph in some sort of persistent store. Obviously, this is going to be a much weirder database than Option 1 gives you; but on the other hand, this approach naturally gives you much more fine-grained change propagation behaviour. That means you don’t have to worry about free variables making stuff confusing. More importantly, it is fully automatable. You just feed your program into the madhouse, and off it goes.

Similar AJAXyness to that of Option 1 is imaginable, but vastly more intricate: you must ship the portions of the dependency graph relevant to the client’s computation down to it. Determining in advance which portions will be relevant is quite likely to be vulnerable to the halting problem, so you’ll have to either send too little (and have a lazy loading strategy to get other data that you need on the client) or too much (and clog the intertubes, OMG).

This option is closest to the approach taken by Asana’s Luna and by Meteor.

Parenthetically: I was amused to discover that with Option 1, the efficient thing to do is to cons a new input onto the head of the list. With this option, the efficient thing to do is to replace the empty list at the end of the input list with a singleton list containing the new input.

Conclusion

I’m sticking with Option 1 for now. I’m about to attempt integrating with a database; if that goes badly, Option 3 is going to start looking much better. Again.

Not In Detention

They’re not in detention, honest! It’s right there in the law (clause 11):

An offshore entry person who is being dealt with under subsection (3) is taken not to be in immigration detention (as defined in subsection 5(1)).

So that’s nice. They’re not in detention! What are they? Well, that’s the question, innit?

They are simply enjoying the hospitality of the Australian government on some island a long way from anywhere, perhaps with no way to leave unless they tell the aforementioned Australian government to send them home.

I say “perhaps”, because there’s no information about what the conditions in these processing places will be. I was looking at the law because I wanted to know what we were doing to protect these people who come to us and ask: please, protect us? I couldn’t find anything, and was hesitant to write about it because I’m not really qualified to read these things. But then I found this. Take us to the bridge, Sarah Hansen-Young:

The Houston report included many important recommendations, none of which this government or the opposition have said any word of since Monday. Where is the commitment to increase our refugee intake immediately? Where is the plan to resettle those who have been waiting for years in Indonesia? Where in this legislation is our commitment to upholding our commitment under the convention to treat people properly, to ensure that they are housed appropriately, to ensure that they have access to legal assistance and to ensure that we do not detain children? None of that stuff, none of those important safeguards which are listed in the Houston report, are in this legislation. The government say that they are implementing the Houston report and they have not even bothered to read the detail of the recommendations.

Bear in mind that I’m not just believing her, I have read the documents. She speaks truth. The legislation enables the first and worst recommendation, offshore deten^h^h^h^h^hprocessing, and government and parliament have done nothing to ensure conditions in those places are tolerable. Nor have they done anything with the report’s suggestions to improve the queue that asylum seekers are supposedly jumping.

Nothing about Houston Report 3.46’s appropriate accommodation, or its physical and mental health services. Nothing about its educational and vocational training, its monitoring arrangements and case management assistance. And most worryingly, nothing about its application assistance or its appeal mechanism.

Nothing about the Report’s 3.22’s increase in places for folks currently in Indonesia, or its “practical agenda of initiatives”, or 3.26’s capacity-building of UNHCR processes in the region.

Nothing.

Please, please, prove me wrong.

But enough of the heavy stuff! Let’s have some comic relief. From Migration Legislation Amendment (Regional Processing and Other Measures) Bill 2012, aka the Stop The Boats bill:

198AC   Documents to be laid before Parliament

(1)  This section applies if the Minister designates a country to be a regional processing country under subsection 198AB(1).

(2)  The Minister must cause to be laid before each House of the Parliament:

(a)  a copy of the designation; and

(b)  a statement of the Minister’s reasons for thinking it is in the national interest to designate the country to be a regional processing country, referring in particular to any assurances of a kind referred to in paragraph 198AB(3)(a) that have been given by the country; and

(c)  a copy of any written agreement between Australia and the country relating to the taking of persons to the country; and

(d)  a statement about the Minister’s consultations with the Office of the United Nations High Commissioner for Refugees in relation to the designation, including the nature of those consultations; and

(e)  a summary of any advice received from that Office in relation to the designation; and

(f)  a statement about any arrangements that are in place, or are to be put in place, in the country for the treatment of persons taken to the country.

In other words, there’s a whole lot of stuff that will help the parliament to make the judgement about if it’s a good idea to send refugees to this country. The minister must provide it. Take particular note of that word “Must”, the third word in (2). Onwards:

             (3)  The Minister must comply with subsection (2) within 2 sitting days of each House of the Parliament after the day on which the designation is made.

What? What are you talking about? 198AB said that if the document wasn’t explicitly rejected by either house within five days, it would be accepted. I’m not clear on how “within 2 sitting days” is counted by these folks, but if there’s scope for the minister to provide the documents at the end of day 2 out of 5, that leaves 3 days for the houses to accept or reject. They’ve built a stalling tactic right in to the legislation! But that’s okay, surely it can’t get any worse…

             (4)  The sole purpose of laying the documents referred to in subsection (2) before the Parliament is to inform the Parliament of the matters referred to in the documents and nothing in the documents affects the validity of the designation. Similarly, the fact that some or all of those documents do not exist does not affect the validity of the designation.

The first sentence says it’s to help you decide about the designation, it doesn’t define or otherwise affect the designation (in network protocol standards, we’d say it’s not normative). Fair enough. But the second sentence seems to come awfully close to giving the Minister a pass on justifying the designation of a country. I thought it couldn’t get any worse? Surely now…

             (5)  A failure to comply with this section does not affect the validity of the designation.

Ah, this one is simple. The word ‘must’ in clause (2) was for cosmetic purposes only. The Minister can not only stall (as per (3)) or partially comply (as explicitly provided for by (4)), he can in fact skip the whole bit.

See? Comic relief, just like I promised. It must be funny, because the primal law of “if you don’t laugh, you’ll cry” is in effect. It’s in effect so hard. Still, at least it’s better than the Morrison amendment.

P.S. Big ups to the folks maintaining the Australian Parliament House website, by the way, you guys rock my world. Austlii’s consolidated legislation listings rate a mention here too, it’s good stuff. If you think the law is good stuff, that is.

P.P.S There is a more charitable reading of 198AC. The pre-amendment rules are that the minister gets to designate places of off-shore processing; in the worst case, the new rules degrade back to that situation.

The Asylum Seeker Trade-off

I wrote earlier that the trade-off between deaths at sea and possible trauma in detention at Nauru could potentially be justified. That’s predicated on a pretty big assumption, the assumption that there are only two options to choose between. Rather than a policy that may encourage people to come by borderline unseaworthy boat, or a policy of deterrence, how about a policy where safe transport from Indonesia to Australia is provided by the Australian government? This policy is often mentioned, but usually with scorn: “what are we supposed to do, just charter a plane to bring them here?” I am interested in actual reasons not to do this.

The Houston Report doesn’t really answer my question, although it provides a hint. Buried in the document, deep within the sweaty groin of Attachment 6: “Australia’s International and Regional Engagement on Irregular Movement and International Protection”, is this note about practical arrangements developed under the Regional Cooperation Framework of the Bali Process on People Smuggling (etc etc): “Any arrangements should avoid creating pull factors to, or within, the region.” Creating a safe and effective mechanism for people who would otherwise be Irregular Migrant Arrivals would create a big-ol’ pull factor to Indonesia, and Indonesia doesn’t want that.

That’s all reading between the lines. The report doesn’t really talk about the possibility. But – should my reading be accurate – Indonesia’s distaste for the notion would probably not be absolute. Could they be persuaded?

Limits

The Houston Report advises re-opening Nauru and Manus Island as off-shore processing centres for asylum seekers. From my point of view, that sucks. I try to be all love-your-neighbour, yo, and shipping your neighbour off to a tiny island – where they are to stay for an unknown period of time – is not real loving. How did such a sucky thing come to pass?

There were three people on the panel that produced the report: Angus Houston, a retired air force man; Professor Michael L’Estrange, an academic and public servant with a rightish bent; and Paris Aristotle, a man with the most awesome name ever but also the Director of the Victorian Foundation for Survivors of Torture. Mr Aristotle’s schtick is supporting and advocating for refugees.

When asked about the possibility of trauma in the new Nauru processing centre, Mr Aristotle compares the potential trauma of prolonged presence on Nauru to that of losing relatives at sea. Given that the rate of boats being lost at sea en route to Christmas Island has increased recently, this issue of asylum seekers dying at sea is a real concern, and the trade-off is potentially justified.

Alas, alack, we can explain all this without considering the trade-off. The panel was assembled to answer a question, and that question takes the form of their Terms of Reference for the report. The question they were asked, Numero Uno, their raison d’être: “how best to prevent asylum seekers risking their lives by travelling to Australia by boat”. Or, as Tony Abbott would put it: “Stop the Boats”. There were other things they were asked to write about:

  • source, transit and destination country aspects of irregular migration
  • relevant international obligations
  • the development of an inter-related set of proposals in support of asylum seeker issues, given Australia’s right to maintain it’s borders
  • short, medium and long term approaches to assist in the development of an effective and sustainable approach to asylum seekers
  • the legislative requirements for implementation
  • the order of magnitude of costs of such policy options

None of these things are the physical or mental well-being of asylum seekers. None of these things are about respecting and caring for vulnerable people looking to us in a time of need. These questions were not asked.

They are kind of between the lines. Our “relevant international obligations” include some treaties that bind us to a certain minimal level of not being a dick. Imposing irreparable psychological damage probably isn’t a sustainable approach to asylum seekers. The well-being of asylum seekers might be considered an “asylum seeker issue” for proposals to be in support of. But this is all implicit stuff, it doesn’t rate a mention in the terms of inquiry.

Compare and contrast this with the lead entry, the Stop the Boats priority. This requirement can be read out of the “inter-related set of proposals” requirement, the “effective … approach to asylum seekers” terms, and maybe even the “destination country aspects”. But to leave it hidden away in there would be to allow the report to come up with any answer at all. The answer that might perhaps buy Tony Abbott’s silence on the issue is off-shore processing, and so the question must focus on deterrence. And thus we asked the panel not to advise us on how to humanely and effectively treat asylum seekers in Australia, but on how to stop them coming.

Paris Aristotle never had a choice. He can be as hard-headed as he wants; he is bound by the terms of reference to answer the wrong question. He is bound to give an answer focused on stopping the boats. He may also think it a justified trade-off in the present circumstances, I don’t know. But it doesn’t matter. His job was to answer the question he was asked. He could have affected the details, but the overall thrust of the report? He never had a choice.