In my station about my short talking at CQC, I adverted that the groupoidification plan in physics is based on a couple of simple conceptions ( most research plans are, I say ). The ones I singled out are: province, correspondence, and history. But since conceptions lean to appear simpler if you leave them vague, there are adhered to be nuances here. Recently I 've been considering about the first one, province. What is a province? What is this purportedly simple conception?
Etymology is n't an especially dependable indicant of what a word intends, or even the history of a conception ( words modify significations, and concepts displacement over clip ), but it Holds sometimes interesting to follow. The English word `` province '' comes from
the Latin verb gaze
, intending `` to stand '', whose perfect participle is position
, which is likewise borrowed straightly into English. The Proto-Indoeuropean root sta-
likewise intends `` stand '', which successively comes from this root, but this clip via Germanic ( along with `` standard '' ). Nevertheless, most of the words with this root come via assorted Latin intermediators: province, stable, position, statue, stationary, station
, and likewise substance, understand
and others. The province of things
is sometimes referred to as being `` how things stand '', how they are, the current status. Most of the words based on the sta-
root imply non-motion (i.e. `` stasis '' ). If anything, `` province '' ( like `` position '' ) transports this intension less strongly than most, since the province of matters can alter - but it emphasise how things stand now
and not how they 're altering. From this sense, we likewise get the political significance of `` a province '', a reified version of a term originally intending the political status of a state ( by analogy with Latin faces like position rei publicae
, the `` status of public things '' ).
So, contracting focusing now, the `` province '' of a physical system is the status it Holds in. In different frameworks of aperients, this is drawn in different shipways, but in each example, by the `` status '' we intend something like a complete description of all the facts about the system we can get. But this intends different things in different scenes. So I simply desire to have a look at some of them.
Conceive of these different scenes for physics as being literally `` scenes '' ( but delight pardon the wordplay ) of the turn on a machine. Three of the switches are labelled Thermal, Quantum, and Relativistic. The `` Caloric '' switch varies whether or not we 're speaking about thermodynamics or ordinary mechanics. The `` Quantum '' exchange varies whether we 're speaking about a quantum or classic system.
The `` Relativistic '' switch, which I 'll disregard for this station, stipulates what rather invariability we hold: Galileian for Newton 's aperients; Lorentzian for Special relativity theory; general covariance for Einstein's general theory of relativity. But this gets into kineticses, and `` province '' connote things are, goodly, still - that is, it Holds about kinematics. At the really least, in Relativity, it Holds not canonic what you intend by `` now '', then the definition of a province must include selecting a reference system ( in Sr ), or a Cauchy hypersurface ( in GR ). So allow 's rubric over that for now.
When all these switches are in the `` off '' place, we hold newtonian mechanics. Here, we believe of a province as - at a first grade of estimate, an component of a set
Now, for serious newtonian mechanics, this set will be a symplectic manifold, like the cotan packet 
of some manifold 
This is really a little subtle already, since a point in 
corresponds a aggregation of places and impulses ( or some induction thereof ): that is, we can commence with a infinite of `` still '' constellations, parametrized by the values of some evident measures, but a province
( contrary to what etymology proposes ) too includes momenta drawing how those amounts are modifying with clip ( which, in newtonian mechanics, is a fairly simple impression ).
The Hamiltonian image of the kineticses
of the system so says us: given its province, what will be the acceleration, which we can so apply to figure provinces at future clip. This demands a Hamiltonian, 
, which we consider of as the energy, which can be figured from the province. So, for instance, kinetic plus p.e.: in the example of a speck going in a potency on a line, 
The infinite of provinces can be drawn without much mention to the Hamiltonian, but once we hold 
, we get a flowing on it infinite, transforming old provinces into new provinces with clip.
Now if we turn along the `` Thermic '' exchange, we hold a different impression of province. The standard image for the classic mechanical system is that we may be speaking about a atom, or a couple of atoms, or mayhap a stiff object, locomoting in infinite, perchance open to some restraints. In thermodynamics, we are believing of a statistical ensemble of objects - in the simplest instance, 
indistinguishable objects - and desire to enquire how energy is alloted among them. The standard image is of a box full of gas at some temperature: it Holds full of molecules, each with its ain flight, and they interact through hits and exchange energy and impulse. Instead than tracking the exact places of molecules, in thermodynamics a `` province '' is a distribution
, or more precisely a chance step, on the infinite of such provinces. We make n't presume we cognise the detailed microstate
of the system - the places and momenta of all the atoms in the gas - but justly something about how these are administered among them. This reflects the existent fact that we can justly mensurate things like pressure, temperature,etc. The step is saying us the proportion of molecules with places and impulses in a given ambit.
This is a large difference for something drawn by the same word `` province ''. Even presuming our inherent infinite of `` microstates '' is still the same 
, the province is no more a point. One mode to construe the difference is that here the province is something epistemological. It draws what we cognise about the system, instead than everything about it. The step replies the interrogation: `` given what we cognise, what is the likeliness the system is in microstate Ten? '' for each Tenner Now, naturally, we could take a infinite of all such steps: given our old classic system, it Holds a infinite of functionals on 
So the province can again be seen as an ingredient of a set. But it Holds more natural to hold in position its nature as a step, or, if it Holds nice plenty, as a positive mapping on the infinite of provinces. ( It Holds interesting that this is an object of the same type as the Hamiltonian - this is, intuitively, the footing of what Carlo Rovelli names the `` Thermal Clip Hypothesis '', summarise here
, which is secretly why I desired to write about this theme. But more on it in a ulterior station. For one thing, before I can speak about it, I need to speak about what comes next. )
Now turn forth the `` Caloric '' switch, and consider about the `` Quantum '' switch. Here there are a few points of perspective.
Earlier, we draw a system in footings of a Hilbert infinite, and a province is a vector in a Hilbert infinite
Again, this could be drawn as an constituent of a set, but the complex linear construction is important, so we hold believing of it equally primal to the type of a province. In geometrical quantisation, one ofttimes begins with a classic system with a province infinite like 
, so takes the Hilbert infinite 
, so that a province is ( modulo analysis issues ) essentially a complex-valued mapping on 
This is something like the ( positive real-valued ) step which gives a thermodynamical province, but the reading is slippery. Course, if 
is an 
-space, we can retrieve a chance step, since the square modulus of 
holds finite entire step ( so we can normalise it ). But this is n't decent to depict 
, and the superfluous info of stages locomotes losing. In any example, the chance step no more holds the obvious version of drawing the statistics of a whole ensemble of indistinguishable systems - justly the likeliness of mensurate particular values for one system in the province 
( As a matter of fact, there are assorted no-go theorems getting in the style of a probablity version of 
, though this again affects kineticses - a recurring subject is that it Holds difficult to conclude reasonably about provinces without kineticses ). So despite some similarity, this conception of `` province '' is rattlingly different, and stage
is a primal portion of how it Holds different. I 'll be jiggered if I can state why
, though: most of the `` huh? '' factor out quantum mechanics dwells right somely here.
Another style to depict the province of a quantum system is associated to this chance, though. The scalar product of ( whether we chance it as an -space or not ) gives a mode to speak about statistics of the system under recurrent observations. Observables, which for the classic image are depicted by maps on the province infinite , are now self-adjoint operators on The outlook value for an evident 
in the province is $ \langle \phi | A | \phi \rangle $ ( note that the Dirac notation implicitly employs self-adjointness of 
So the province holds another, intuitively easier, version: it Holds a real-valued functional on observables, viz. the one I only drawn.
The observables sleep in the algebra 
of sprung operators on 
Positioning both Thermal and Quantum switches of our impression of `` province '' gives quantum statistical mechanics. Here, the `` C*-algebra '' ( or von Neumann-algebra ) icon of quantum mechanics states that verily it Holds the algebra 
that Holds key - it matches to existent operations we can do on the system. Some of them ( the self-adjoint ones ) correspond rattlingly rattlingly intuitive things, viz. observables, which are touchable, mensurable amounts. Therein icon, 
is n't presumed to commence with in the least - but when it is, the rather object we 're addressing with is a density matrix
This is ( roughly ) a positive operator on of unit suggestion ). Generally a state on a von Neumann algebra
is a linear functional with unit suggestion.
This is correspondent to the position of a province as a chance step ( positive map with unit entire built-in ) in the classic kingdom: if an discernible is a map on provinces ( giving the value of that evident in each province ), so a step is so a functional on the infinite of observables. A chance step, in point of fact, is the functional giving the outlook value of the evident. ( And, since variant and all the higher moments of the chance distribution for that evident are themselves delimitated as outlook values, it too states us all of those. )
Then again, the Gelfand-Naimark-Segal theorem
states that, given a province 
, there Holds a representation of 
as an algebra of operators on some Hilbert infinite, and a vector 
for which this is only 
This is the GNS representation ( and in point of fact it Holds maked by taking the regular representation of 
on itself by generation, with done into a Hilbert infinite by definining the dot product to do this belongings work, and with 
So the perspective here is that a province is some sort of operation on observables
- a much more epistemological perspective of things. So although the GNS theorem associates this to the vector-in-Hilbert-space perspective of `` province '', they are rather different conceptually. ( For one thing, the GNS representation is giving a different Hilbert infinite for each province, which sabotage the sense that the infinite of ALL provinces is fundamentally `` there '', but in both pictures is the same for all provinces. )
( This von Neumann-algebra point of position, incidentally, gets on nicely with the 2-Hilbert infinite lense for looking at quantum mechanics, which may partially bridges the spread between it and the Hilbert-space perspective. The class of representations of a neumann algebra is a 2-Hilbert infinite. A `` 2-vector '' ( or `` 2-state '', if you care ) therein family is a representation of the algebra. So the GNS representation itself is a `` 2-state ''. This raises the enquiry about 2-algebras of 2-operators, and Lav Baez ' interrogation: `` What is the categorified GNS theorem? '' But permit 's leave 2-states for subsequently along with the balance. )
So where makes this leave us considering the import of `` province ''? The classic perspective is that a province is an factor of some ( structured ) positioned. The usual quantum image is that a province is, dependent on how precise you desire to be, either a vector in a Hilbert infinite, or a 1-d subspace of that Hilbert infinite - that is, a point in the projective Hilbert infinite. What these two positions hold in common is that there is some infinite of all `` possible existences '',i.e. of all slipways things can be in the system being analyzed. A province is so a fashion of taking one of these. The difference is in what this infinite of possible creations is like - that is, which category it sleep in - you said it exactly one `` selects '' a province. How they differ is in the possibility of taking combinations of provinces. As for choosing provinces, 
is a Cartesian class, with a depot object 
: an factor of a set is a map from 
into it. 
is a monoidal class, but not Cartesian: choosing a individual vector holds no obvious categoric equivalent, though taking a 1-D subspace sums to a map from 
( up to isomorphy ). So the framework of an `` factor '' is n't a singleton, it Holds the complex line - and it links to other possible infinite differently: not as a terminal object, but as a monoidal unit. This is a categoric manner of telling how the thought of `` province '' is structurally different.
The caloric point of position is a bit more epistemically elusive: for both classic and quantum pictures, it Holds best idea of as, not a possible macrocosm, but a map moving on observables ( that is, conditions of cognition ). In the classic ikon, this is directly linked to a infinite of possible cosmoses - it Holds a step on that, which we can believe of as stating how a big ensemble of systems are dispense therein infinite. In the quantum ikon, in some slipways the most ( epistemically ) natural perspective, in footings of neumann algebras, interrupts the connexion to this feeling of `` possible cosmoses '' wholly, since holds representations on many different Hilbert infinite?
So a philosophic inquiry is: what make these different constructs hold in common that permits us apply them all to correspond the `` same '' root thought? Without really replying this, I 'll but name that at some point I 'd wish to speak a trifle about `` 2-states '' as 2-vectors, and generally how to categorify everything above.
![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small}
