It is a commonplace to say that the state is divided into three branches: executive, legislature and judiciary. Each, quite rightly and quite deliberately, in a condition of constructive tension with the other.
The controversy over the role of the Supreme Court in Scotland adds a further dimension: which state or, more precisely in this case, which nation?
The Advocate General has tonight entered the joust. No, not the Lord Advocate.
The A-G is Lord Wallace of Tankerness. Jim Wallace, as was, the former leader of the Lib Dems and deputy FM. He is the UK government's law officer who keeps an eye on Scottish matters.
He argues that the "people of Scotland should have their human rights protected to the same level as those in other parts of the UK". This, he believes, is best done via the UK Supreme Court.
Alex Salmond - and those who support his desire to constrain the Supreme Court's role in Scotland - back Lord Wallace's general statement of intent. But not his conclusion.
Let us note a few points. I have heard it said in some quarters that this should not be a constitutional or political debate: that it is somehow mechanistic or organisational.
That is to misunderstand the nature of a state. It is a fundamental element in politics to weigh the balance between the various arms of state power.
Further, it is a fundamental element in Nationalist politics to seek to ensure that those arms of state power are wielded by Scottish hands as far as possible. It is an entirely legitimate debate.
The distinct nature of the Scottish judicial system was stressed in the Act of Union. Clause 19, since you ask, in there behind the provisions on coinage and salt.
However, the Union rules on Scottish justice are, at sundry points, made subject to "such Regulations as shall be made by the Parliament of Great Britain".
One might argue that the current row springs from one of those regulations. Except that it was not meant to be like this.
It was not intended that the UK Supreme Court would arrogate to itself a role in determining Scottish criminal cases.
Which brings us to another point of difference. Supporters of the new system insist that has not happened: that Scots Law remains broadly untouched, that any interference is only to uphold human rights, as envisaged in the 1998 Scotland Act.
They note that only a tiny number of cases affecting Scots Law have been heard by the Supreme Court with many others turned away.
Critics say the intervention goes beyond that: to the extent that the Cadder ruling last year potentially affects thousands of cases and alters the day-to-day operation of Scottish justice.
They say further that, unlike in England, the Scottish High Court has no say on whether Scots criminal cases may go to the Supreme Court.
A few points. Much of this arises from the 1998 Scotland Act which incorporated the provisions of the European Convention of Human Rights into the proposed devolved structure.
From the establishment of devolution, in 1999, it was provided that disputes involving the exercising of devolved powers could go to the House of Lords, to the Judicial Committee of the Privy Council.
One or two droll wags pointed out that the historic role of Committee was to rule on disputes involving UK overseas territories and colonies. Nonetheless, devolution fell under their remit.
On the establishment of the Supreme Court, that role of settling devolution disputes transferred to the new body.
So, when the UK Court rules on Scottish cases, it is not judging the facts per se. It does not, for example, determine guilt or innocence. It does not occasion, of itself, the release of a detained person. It is judging whether the ECHR provisions of devolution have been followed, most notably by the Lord Advocate.
Appeal to Strasbourg
In summary, supporters of that system say it provides a ready interpretation of human rights law: far quicker and more convenient than going to the European Court in Strasbourg.
Critics say it is too wide-ranging: that, by extension, almost anything can be interpreted as affecting the human rights of an individual.
Mr Salmond notes that, in the interim, it should be established that the UK Supreme Court can only intervene in Scots cases on invitation from Scotland.
He argues further that Scotland could establish a distinct bench in the High Court to rule, separately, on such matters. He notes that, in any case, there is always provision for an appeal to Strasbourg, even beyond the UK Supreme Court.
His critics say such a model is unwarranted, unwanted, costly and ineffective by comparison with an established UK court, which has two senior Scottish judges on its strength. They say Mr Salmond is, in effect, discontented with "where" rather than "what" when it comes to assessing the UK court.
So there are arguments about efficiency and effectiveness: Supreme Court v Strasbourg. There are arguments about scope: minimal or maximal intervention. There are arguments about trust: will minimalism prevail or slip? There are arguments about cost and system.
But there is also a fundamental argument about the constitution.