Problems with normative/descriptive claims about law; vol. 1

A basic distinction in legal scholarship lies between descriptive claims (what is the law?) and normative claims (what should be the law?). In this post I argue that one can easily fall into the trap of making one instead of another, due to: 1) law’s normative character; 2) law’s hierarchical (or heterarchical) structure; and 3) often argumentative style of legal reasoning.

Initial claims:
1. Since law itself is normative, a descriptive claim about law will be a normative claim about sth else;
2. A descriptive claim about a higher order provision (in Kelsen’s sense or some heterarchical sense, e.g. about the EU law) will be a normative claim about a lower order provision;
3. When law is ambiguous, a seemingly descriptive, ‘interpretative’ claim might actually include extra-legal normative components.

Examples:
Assume a hypothetical state, the Republic of Silesia, where the Traffic Regulation Act sets the default speed limit at 60 km/h in the towns.

Descriptive question: what is the default speed limit in the towns in the Republic of Silesia? is pretty straightforward, and the answer: ‘the speed limit is 60 km/h’ is a correct, descriptive claim.
But this claim already includes a normative statement: ‘people should not drive faster than 60 km/h in Silesia’.

However, a normative question: what should be the default speed limit in towns in Silesia? invites different possible types of answers:
#1: The speed limit should be lower, because I’m afraid of cars driving so fast;
#2: The speed limit should be higher, so that people can be more efficient, move faster and so have more time to work, due to saving time on transportation;
#3: The speed limit should be higher, because the art. X of the Constitution of the Republic of Silesia states ‘The Republic ensures the citizens’ right to move around quickly’.

Higher/lower limit is insignificant here, what matters is the reasons behind the normative claims.
In #1, this is just an opinion;
in #2 a higher-level normative claim: ‘law should maximise the efficiency’ is assumed, though this claim is external to the legal system;
in #3 a descriptive claim about higher level law (constitution) is made, together with a normative claim: ‘lower level norms should comply with higher level norms’, which is internal to the legal system, and could be summarised as an internal normative claim: ‘law should make sure that citizens can move around quickly’.

Note that #2’s normativity comes from outside of the legal system, while #3’s from the inside.
Why does it matter?
Because one could claim that a statement: ‘The speed limit in towns in Silesia must be higher, because the current one is unconstitutional’ is actually a descriptive claim about the law.
In other words: seemingly normative statements about the law (due to the lingual side and usage of the modal verbs like ‘should’, ‘ought to’ or ‘must’) can be considered descriptive statements by some scholars.

One should bare in mind that when making any claim about the law, one sees at least 3 levels:
1. Reality;
2. Legal provisions in statutes;
3. Constitutional/human rights provisions.

When making a descriptive claim about a higher level, one makes a normative claim about the lower level; when saying what the law is, one says how the reality should be; when saying what the constitutional norms are, one says how the lower level law should be. All these normative statements are internal to legal system.

However, there is a space for ‘cheating’ here. For example, it is not obvious that the speeding limit must be higher, maybe it is sufficient. This is already an argumentative exercise.

Now, the picture gets more complicated when one makes a normative claim based on reasons external to the legal system. These claims, again, can be directed at different levels:
1. People should be moving around quickly to be efficient;
2. Law should protect safety of the citizens, because human life is the highest value.
3. The silly provision about the right to move around quickly should be taken out of the Constitution, because it only causes confusion.

The importance of the distinction is high, because one can argue that executing a normative proposition internal to the law is a legal action, while executing one external to the law is just political action.

Now, this gets even more complicated when legal provisions are not straightforward.
Assume the Constitution states: ‘The Parliament of Silesia elects the Justices of the Constitutional Court and President of the Republic accepts the vow from them’ and assume that the Parliament has elected 4 justices, but the President refused to accept the vow.
The question begging a descriptive answer: ‘Is the President allowed not to accept a vow from the justices?’ needs an argumentative argument actually; and the reasons invoked might be internal, but might just as well be external to the legal system.
One can imagine both an affirmative and a negative answer, about which it is hard to state whether they describe the law, or just state as it should be.

In consequence, a rigid scholar should be always be explicit about the internality or externality of the source of normative claims he or she is making; and about the level about which a claim is made. Otherwise, falsity of the theories might hide to deep in the argument to be detected.

The trouble with ‘What is X?’ questions

I’ve just solved a problem which has bothered me for two years, and in my excitement want to share the solution with the world.

The problem, in short, is: ‘What is X?’ questions (‘What is law?’, ‘What is religion?’, ‘What is mind?) are in English necessarily ambiguous in at least one sense. They can be questions about both:
1) the delineation of the scope of the inquiry (out of all the phenomena in the universe {x1, x2, x3, …, xn}, which phenomena are to be denoted by the term ‘X’ in a particular inquiry); and
2) the insight into the object of inquiry (having agreed that ‘X’ means {x3, x17, x56 and x875}, how is one to understand that phenomenon).
This seems obvious, but scholars surprisingly often get that mixed, and in consequence talk past each other (details below).
Why? Because English language does not differentiate between cases.

Now, my problem was the inability to communicate that, in other words, if you have not understood what I’ve just written, don’t worry, it’s my fault.

Further, this problem would not have occurred in any language with a robust system of noun declination, and by robust I mean one that uses the instrumental case, like Czech, Russian or Polish. In the last one, ‘what is law?’ question would be translated as ‘Co jest prawem?’ in the first meaning (delineation) and ‘Czym jest prawo?’ in the second (insight). Giving an answer to one while asking another would just feel obviously illogical due to the grammatical structure of the sentence. Hence, whenever talking about the problem above to someone speaking a slavic language, I got this ‘aaaaaaaaaahhhhh!’ when just translating the sentence.

Unfortunately, none of the Western languages that even use cases (German or Latin), so none of the languages that I could have expected majority of my interlocutors to understand, use the instrumental case. So I couldn’t express what I mean.

And now I got it. Two different meanings of ‘what is X?’ question can be exposed in English via the sentence structure of a negation of the question, as in:

what is not law? (negative delineation) – Co nie jest prawem?
what is law not? (negative insight) Czym nie jest prawo?

So, if you agree that these questions are not the same, you can see how a non-negated question ‘What is X?’ can always mean at least two different things.

I am super excited, I don’t now how about you.

Why would that matter? Take the 25-centuries-old, unsolved question: what is law?

I would claim that the trouble with this question lies not only with the object (‘law’), but also with the method(s) and with the question itself; one of the problems with the question being the one explained above.

Take H.L.A. Hart, who in the most overrated law book of the 20th century, The Concept of Law, starts by identifying the diversity of answers:
1) What officials do about disputes is … the law itself (Llewellyn)
2) The prophecies of what the courts will do . . . are what I mean by the law (Holmes)
3) Law is the primary norm which stipulates the sanction (Kelsen)
Clearly, ‘officials’ actions’ are not ‘prophecies’ and are not ‘norms’. The three answers above are answers to the first meaning of the question, just delineations of scope, different ones.

Hart however, to a great surprise of a reader, just skips this problem and implicitly answers the first question by limiting ‘law’ to norms, and then proceeds to the second one, to which he gives the famous answer about the union of the primary and secondary rules.
How come?! I seriously think he just didn’t notice, because he was writing in English.

So, to finish, let us negate a few more questions, just to get the feeling:

What is religion?
1. what is not religion?
2. what is religion not?
What is mind?
1. what is not mind?
2. what is mind not?
What is love?
1. what is not love?
2. what is love not?

CREDITS: Great thanks go to Józef Życiński, over whose book Elementy filozofii nauki I got the idea. The book is in Polish, I was taking notes in English, and the hesitation over the sentence structure of ‘What is not a scientific theory?’ was the eureka moment for me.
Special thanks go to Bartosz Marciniak, who spend a significant amount of time and characters to show me how to translate the two into Latin, finally explaining to me why this could not be expressed in Latin the way I wanted to do it.
Additionally, huge thanks go to Liam McHugh-Russell, who quickly spotted a problem that could have ruined the whole post, now fixed already.

Anthropology before philosophy (?)

The question I want to consider today is: Must one do anthropology before one does philosophy?

One of the fields of legal scholarship that would benefit from an answer is the now-hip field of the Transnational Legal Theory. The question could be specified as: If one wants to propose a theory explaining the phenomenon of the transnational law, must one first empirically examine the phenomenon in question?

And a tempting answer is: (hopefully) not. Hopefully, for philosophers don’t like to get their hands dirty with empirical research. And a compelling argument could be offered in defence of that answer: none of Kelsen, Hart, Dworkin, or Raz ever did any anthropology, and that their theories seem to do just fine without it. And even if the theories themselves aren’t fine, the objection is not “it seems like you guys didn’t do enough anthropology.”

Let me moot two claims.

The answer to the first question is it depends: sometimes one needs to do anthropology before one does philosophy, sometimes not. In particular, it depends on a) existence of ‘folk’ intuitions about the phenomenon to be theorized; b) these intuitions actually being true (corresponding to facts).
The answer to the second question is yes: to provide a theory of transnational law, one first needs to engage in some anthropology (or at least base the theory on the work of others who have). That is because theoreticians and/or their audience still lack sufficiently developed intuitions about the phenomenon of transnational law. Worse, the intuitions they do have often happen not to be true. The context therefore departs significantly from the intellectual milieu in which Kelsen and Hart did their theorizing about ‘traditional’ national and international law.

Let me take a moment to clear up some terminology. ‘Anthropology’ here means, very roughly, empirical research on reality (talking to people, observing phenomena, reading texts) i.e. the any collection of the data that can’t be generated just by sitting in armchair and thinking. ‘Transnational law’ means norms generated and executed by entities that are neither national nor international, ranging from the EU and the WTO, to FIFA and UEFA.

Example 1: Information vs. BitCoin
Start by considering an example of two research questions, with superficially similar structures: what is information? And: what is a cryptocurrency, like BitCoin?

One can easily imagine providing an answer to the first question via armchair philosophy. A researcher would sketch some working ontology, differentiating a piece of information as either: 1) a feature of an object (e.g. a fact of me being a Polish citizen); or 2) an expression of that fact (Filipe saying: ‘Przemek is a Polish citizen’); 3) a fixation of that expression in some context (EUI officer writing in the database ‘Przemek is a Polish citizen’) etc. He or she might then theorize the differences between these types, suggest an ontological explanation, and potentially arrive at some definitions. All this is possible because the researcher ‘knows enough facts.’

What is BitCoin, however, will be harder. For before one can offer a philosophical explanation of its existence and functioning, one needs to know what BitCoin actually is. One needs to obtain the facts. If, for example, one starts by assuming that BitCoin is a private currency, ‘an alternative mode of payment provided by a private company’, and theorizes that, one is simply wrong. For there is no company, and one would completely miss the ‘decentralized’ structure of a crypto-currency. Researcher needs the facts first.

Example two: Philosophy of the Mind
Take another question: what is the relation between a brain (body) and a mind? One could attempt to answer this controversial question solely through ‘armchair philosophy’; in the end, we have quite good intuitions about what brain and mind are. But in doing so, one could simply be wrong about the facts. The booming career of neuroscience shows that our intuitions about how brain functions had often been mistaken. Now, to be clear, I am not saying that neuroscience will ever replace philosophy of mind. I am saying that neuroscience informs it. It provides facts about which we can theorize. It’s the anthropology necessary before philosophy.

Legal theory vs. Theory of transnational law

Now let us apply these findings to legal theory and theory of transnational law.

‘Hart and Dworkin never did any anthropology’. Well, didn’t they? They both studied law and worked as lawyers before they undertook their philosophical projects. They were intimately familiar with phenomenon of ‘law’ that they were addressing, from both a theoretical and a practical perspective. And their audience was people who shared that intimate knowledge as well. They had the facts right.

Imagine someone who had never studied law trying to provide a new, descriptive theory explaining how law functions. There is a chance it would be brilliant and original, granted. But there is a much higher chance that would be wrong, because ‘that is not what actually’s going on here, pal’.

With transnational law, most scholars are in just such a position. We hardly study the phenomena at school. Many of those engaged in legal practices never deal with transnational law directly. We know about some of the relevant institutions, like UE or FIFA, but there are many others we may have never heard about. And we might be just mistaken in our intuitions about them.

A successful theory of transnational law needs to be based on observed facts. Otherwise, theorization will provide an insight not into its desired object, but into the mistaken idea that the researchers have of that object. And so the anthropology needs to come first.

Ontological vs. methodological assumptions

It will take you 6 minutes to read this post

There is a significant difference between methodological and ontological assumptions.

One of the prerequisites of a good scholarship is ability to draw distinctions. To see critical differences between one thing, subject, method, process (…) and another. Especially if the two seem very similar or even the same. What has started with Plato’s critique of Parmenides’s confusion over the meaning of a verb ‘to be’, continues till now as a feature of what Westerners label philosophy, or to be more humble, scholarship. The aim of this post is to draw attention to the one signalled in the lead – between assumptions about ontology of the object of inquiry and assumptions internal to the method of inquiry.

This post is inspired by one sentence uttered by prof. Bratosz Brożek about a year ago, over coffee and a cookie, during Law & Logic summer school. We discussed law & neuroscience, a hip subject nowadays. When I mentioned a problem that I’ve considered an important one in philosophy – the problem of reductionism (assumption that mind can be reduced to physical, chemical and biological processes occurring in a brain), prof. Brożek just looked at me, quite annoyed, and said:

‘Those people (reductionists) confuse methodological assumptions with ontological ones‘

That was quite illuminating. So I’ve been thinking about it for a year now. I wanted to write an article about it, but of course found no time, and so hope that a blogpost will suffice for the time being. To be clear – I am not referring Brożek’s position here – I actually don’t know what he really meant. I just had to share my state of thought on the statement, which I personally consider extremely important.

I proceed in three parts. Firstly, I try to generalise, to explain what is the difference between ontological and methodological assumption in general. Secondly, I apply it the the field of neuroscience & law/philosophy/(…) and the brain/mind problem. Thirdly, I apply the distinction to ‘what is law?’ debate, especially to the natural law vs. positivism sub-debate.

What should go without saying, but let me say it just in case – there is probably tons of literature on these subjects, which I do not refer to, because I have not read it. That is why this a brief(ish) blogpost. Not a scholarly article.

I. In general

Every scholarly inquiry assumes a particular reality. It delineates its fragment, a fragment that it wishes to describe, explain and (possibly) criticise. It asks a question and sets up a method of answering it. Here the source of the distinction lies. For a method is a function of a question, and so the assumptions internal to the method will be a function of the question, not a function of the reality itself.

Take the example of the evolutionary theory of the origins of species and the problem of God’s role in the creation. In order to explain evolution, within the positivist paradigm (what we nowadays limit ‘science’ to), one does not need to assume existence of God. Probably, it’s even better not to do so. Each good scientist shaves with the Ockham’s razor, and so an evolutionist should be able to explain the origin of species relying just on the phenomena ‘matter-up’. Explanation not involving God will be a better one. This, however, does not yet mean that there is no God. This just means that assuming God is not necessary in order to answer the question. But God might be out there, creating the World through evolution. And He might be not. This does not matter. And the other way round – If the question would be ‘can the story of Genesis be reconciled with the theory of evolution?’, then existence of God would be a possible methodological assumption, even if the scholar would personally believe that He does not exist.

II. Neuroscience and X

The problem I signalled above – the problem of reductionism – could be summarised in having to accept or reject the claim: ‘The Mind, meaning the sum of intentional, intellectual, emotive etc. mental states can be reduced, and so described and explained, by solely ‘natural’ phenomena of physics, biology and chemistry’. Philosophically it is hard to defend, and yet many neuroscientists, often seeming as smart people otherwise, defend it vigorously. I claim, after Brożek, that this perplexing state of academic affairs can be blamed on the confusion of methodological assumptions with ontological ones.

For, in a sense, neuroscience needs to be reductionist in its method. Unless the assumption is there, the results would need to be always preceded by a word ‘probably’. Unless the mental states can be explained by the analysis of the brain, the analysis of the brain in order to explain the mental states would be silly. So we assume it is.

This, however, does not mean that ontologically speaking, mind is brain. There are many philosophical arguments against this claim. But as in the example of evolution, one does not really matter for another. The problem appears when a neuroscientist makes a claim that his or her methodological assumption is an ontological assumption. The argument: ‘Since I need to assume X for my method to make sense, and it seems that sometimes my method leads to true conclusions, then X is necessarily the case as a feature of the reality’ is logically invalid.

III. Natural law vs. positivism

‘What is law?’ question should trouble any lower with at least vague interest in philosophy and theory. I do not aim at answering it here, of course. But I am in the middle of a project of explain why it remains so hard to answer. In this post I just point to one of the reasons, being exactly the confusion of a method with ontology.

For the purpose of this post, take ‘legal positivism’ to mean ‘an approach according to which all the law is man-made, and if something is law, its pedigree can be traced down to sources acceptable in a given society’ and ‘natural law’ mean ‘an approach according to which, apart from man-made law, there is some sort of higher law, given by God or discoverable in nature, unchangeable by men’.

One can easily imagine a descriptive research question concerned just with human practices, e.g. ‘what is the law concerning drinking alcohol in public in Poland, and what is the official practice of its enforcement?’. In order to answer it, one would hold first doctrinal and then empirical legal research, taking social practices as sources and disregarding any assumptions about higher law. This, however, does not in itself mean that there is no natural law. On the other hand, one could imagine a question ‘what are the possible ways of regulating in-vitro fertilisation, such that they would not be contrary to the teaching of Catholic Church?’. If the question is so, since CC assumes the existence of natural law (in ontological sense), the researcher would need to methodologically assume it as well (in order to reconstruct it). But this does not mean that any natural law is out there.

Ontologically speaking, natural law exists or it does not, though it’s probably hard to find out whether it does. But this has no significance for legal method. For the method, and its assumptions, will depend on the question asked.

Scholarship is aimed at a truth, and this truth (to follow Kuhn), will always be relative to the method. It has little to do with the Truth, unless it claims to do. But this claim, one way or another, need to be explicit.

Otherwise, people will just talk past each other.

Red card and/or jail? or when is a foul still legal

It will take you 4,5 minutes to read this post

Coppa Pavone, EUI soccer cup, starts tomorrow. Good time to share a few thoughts on intersection of sport rules and criminal/tort law.

The question of this post I came up with about 3 years ago, together with my friend Kazik, when we were looking for a subject to write a master thesis about. None of us chose it in the end, but I still find it challenging, both substantially and methodologically. Feel free to pick it up.

Question: where to draw a borderline between lawfulness and unlawfulness of an action breaking sport rules? In other words: when is a foul still legal, and when is it illegal?

Substance

It is often the case that an action which would normally be unlawful gets legalised through the consent to the sport rules/ sport risk. Consider two examples: boxing and soccer.

Soccer is a sport with a risk of injury, though unlike boxing it is not about injuring the opponent. Assume three situations, two easy, one hard, when one player kicks the other one and breaks his ankle.
1) The player acts in accordance with the rules, e.g. when tackling kicks the ball first, and so there is injury but no foul. In this case court action won’t be successful, for the injured one agreed to the rules and the injurer complied.
2) Another player approaches the injured one, lying on the ground, and starts kicking him on the other ankle, breaking it. He clearly breaks the rules of the game, there is a foul, and clearly there is a tort and a crime, for his actions are in no way legalised by the rules.
3) But consider an action in which, while fighting for a ball, one player kicks another one in the ankle and unintentionally breaks it. He breaks the rules, but very often all would agree that he is not liable for tort and guilty of crime – such things just happen in soccer. If every foul resulting in injury would be illegal, no one would dare to play the game. So the consent seems to justify not only the actions complying with the rules of the game, but also those crossing them and… exactly, and what? just the intention seems to simple.

One might seek a borderline – ‘up to this point there is not illegality, right above it there is’ – or one might try to develop a test, not being a clear line itself, but allowing to deem an action lawful or unlawful in every given situation. The other one seems more plausible. But how to do it?

Boxing is easier case in one aspect, and harder in another. Since its very purpose is harming the other player in accordance with the rules, the border of breaking the law will be much closer to breaking the rules. Though still, this won’t always be the case. Still, it is harder, for… how come this is legal at all? For even if me and you agree that we will fight each other outside of the bar, and we agree on rules, if one gets injured, this will be a crime and most probably a tort.

What is so special about the sports?
What is the role of international sport associations setting up the rules?
Are they prior to the state and just accepted, or do they have any sort of delegated power?
How to explain this through legal theory?

Here again, sports might be professional and then it’s easier, but while playing football on a pitch of Villa Schifanoia we are hardly professional, rather nobody reads rules of FIFIA, and yet we come within the sport exception. But if we do box there, we don’t.

Method

This post is not about methodology, but let me just point to a few issues here, for it is a good example of how one can get a legal method wrong. The question: when is a foul still legal, and when is it illegal? makes a bad research question, unless we specify what is meant by ‘being legal’ in the given context. There is a dialectical relation between question and method; and each method has got its limitations when the ‘amount of truth’ it can bring is concerned.

The question might be rephrased as one of the following:
1) where do courts and adjudicators draw the line? then research is empirical and descriptive;
2) given the text of law as it is, how should the test be constructed? then work is doctrinal and seemingly descriptive of law, though unless there is a clear provision, the line between description and prescription gets blurred; still the normative statement here will have to be based on the normativity ‘internal’ to the law;
3) given the rationale of law and social ordering in given society, how should the existing law be interpreted? then research is normative, in the ‘external’ sense, but still bound by the provisions, since considered with interpretation and not a change;
4) given the rationale of X (here legal system in total, cultural setting, social purposes), what should be the test? then research is normative, external and unbound, but… it is very easy to disguise it as still a doctrinal research, and unfortunately this often happens in legal scholarship.

With this let me finish, and wish everyone many goals and zero injuries since tomorrow onwards!

Forza Ladybirds!

Must legal provisions be textual?

It will take you 3,5 minutes to read this post

Is textuality a necessary feature of legal provisions?

A while has passed, but my excitement still keeps high, since the InfoSoc Workshop on Contract Visualisation, paired with Legal Design Jam, offered by wonderful Stefania Passera, Concept Designer & Visualizer, PhD Researcher at the Aalto Univeristy. In this short post I won’t be summarising her argument, but share the thoughts that have buzzed in my head for a week now.

The problem is: contracts are long, people don’t get them (in B2C) or think they agreed are later realise they didn’t (in B2B). Text as a tool is good, but far from perfect. To quote Passera here: ‘contracts often fail not due to legal mistakes, but communication mistakes‘. Proposed solution: let’s put pictures inside, and more generally: pay more attention to visualisation and design.

Example: passage of risk in shipping contracts (from Stefania Passera):

Question that bothers me: must contract (and actually law as well) be textual, or could it be more varied in form? In other words: is textuality a necessary feature of legal provisions; or are they embodied in text because that’s the form we historically got used to? A different set of questions, which I will not consider but state here, is on the very bottom of the post.

It is worth realising that in many instances practices of supplementing, though not yet replacing, text with pictures, in order to facilitate communication, already take place. Let me here point just to two: Creative Commons licenses and App Licences in Google Play:

However, there is a significant difference between representation of existing and textual law in a different form, and creating independent non-textual legal provisions. About a year ago, during GRIM Winter School in Stockholm, prof. Peter Dahlgren, speaking about visualisation of legal material, claimed that trade-off here is between clarity of communication and precision, with pictures being clear but imprecise, logical formulas precise but unclear, and text somewhere in the middle. So it’s best to use all to represent.

But could we use all those forms to make law?

In Polish legal theory we cherish the distinction between a legal provision (a textual unit) and a legal norm (a non-textual sollen, derived from a legal text, often from many provisions, and needing language to be stated, but not being a lingual statement in itself).
If one accepts this distinction as ontologically possible, the conclusion that follows is: law as statutory law must not necessarily be textual; if there is better form, form better realising meta objectives of stating the law, then the better norm should be chosen.

The obvious examples that come to my mind here are traffic signs or non-smoking or no dogs signs (a crossed cigarette or a dog), first being examples of post-statutory regulation, second of either communication of public law, or of contract law (if the owner of a cafe decides not to allow dogs inside). The first, however, are first explained by text in the statue.

Could one imagine a judge interpreting a picture contract clause on passing of risk (like in the first figure above) or a crossed car at the entrance to the park (to use Hart’s classical non-vehicles example?). I could. But would this interpretation be text-free, or would judge first need to textualise the picture as well? But even if so, does this mean that law needs to be text?

To my mind, statutory law or ‘written’ contract do not necessarily need to be represented in text. But I’m happy to hear any criticism of this view.

Other questions here are:
1) When to replace text by pictures, when to pair one with another, and when to draft contracts in technical jargon and then separately create info-graphics to just visualise them?
2) When could it be useful to add pictures to a contract?;
3) Should it be obligatory for entrepreneurs to do so in consumer contracts? (like it is now, in some places, to e.g. draw sugar cubes on Coca-Cola cans);
4) Could using pictures help contract drafters in realising agreement’s weak points? Is there a parallel between conversion to a picture and conversion to XML, when need to clarify meaning of provisions and pictures are concerned?

All these seem like good questions for a paper to me. Feel free to take them!