arXiv:2007.05300v2 [physics.hist-ph] 13 Jul 2020 omto:a gn a ikoeamong one pick in- can generates agent choice an the choose and formation: can in- action agent are of an asymmetry, courses physics Third, alternative time of reversal? the time laws of under elementary origin variant the the that is future, what considering the affects past: not agency the do Second, not that nature. systems of the physical laws violate themselves of that given are independence choices, agents alternative real of assumed possibility its the and agent understand to how oacutfr o oasm.Teol ibealterna- viable only The want assume. we to what is not this for, (as because account perspective contexts), to agent’s many the in of done orientation is time the to cur freedom, cap- or agent. These independence, the to used. the attribute and of we intended degrees which is different in agency ture manners of distinct notion distinguish the also of but bit agency, a here. requires done is that in what something for is is accounted This be- laws be thinking. actual these can the of agency how terms denote understanding we that but known haviour with nature; conflicts of agency de- laws in [ independently nothing in that Loewer been Barry understood has by recently here very considered veloped one the to [ itself physics of foundations the even generating o ntnei icsigtecneta ai fthermo- of basis [ theology. conceptual contexts, dynamics the to foundational discussing in economy in in instance from used Agents utilised for spacing increasingly is areas are meanings. agency in They variable of role with notion a contexts, The play of variety it. a affect and world, nvrosohrrcn hsclprpcie nagency, on [ perspectives physical see recent other various On from? originate information this gnyrie he usin o hscs.First, physicist. a for questions three raises Agency gnyi h osblt o naetto agent an for possibility the is Agency oacutfraec’ ieaymtyw antre- cannot we asymmetry time agency’s for account To of characterisation general a consider I end, this To eeIcnie ouin oteetrequestions. three these to solutions a consider I Here 9 – 12 n eeecsteen esetv similar perspective A therein. references and ] 1 I , hsmyb h eea ehns ttesuc ftewoein whole the of source the tran at that mechanism mechanism general the physical be t a informa may its the therefore This trace of is agency, n source Agency I model the entropy. to identify cases. us and different irreversibility, allows in cal that differ memory, differen above and distinguish questions agency I the agenc to closed. that answer observe appears the o stand I dynamics time the action. agency’s which from an for under agency choosing reason in of the generated notion (ii) mation agent; the an of system aspects physical three Lond discuss N I St. Richmond 1151 Philosophy, of Institute Rotman The log = 2 ,qatmmcais[ mechanics quantum ], .TEPROBLEM THE I. i asil nvriy nvri´ eTuo,CR,CPT CNRS, Toulon, Universit´e de University, Marseille Aix eiee nttt,3 aoieSre ot,Wtro,O Waterloo, North, Street Caroline 31 Institute, Perimeter 2 N iso nomto.Weedoes Where information. of bits 3 8 ,cuaiy[ causality ], ]. N 13 alternatives, .Igv for give I ]. act gnyi Physics in Agency Dtd uy1,2020) 14, July (Dated: nthe on al Rovelli Carlo 4 – 7 ], ohn lea u ipsl[ is be disposal there must our because at agency else principle of nothing second orientation the time in rooted The ultimately past. the in low ut nterlto ewe eoyadetoy[ entropy and memory between re- relation recent a the and on memory, sults and agency between similarity agency. drive model can simple growth a entropy using how filled-in, illustrates be that can agency and namics etl omlto fbt fteetere st see to is theories these of both of theory. formulation quantum or fertile thermodynamics A of foundations the etepiaysuc fteifraintebiosphere, with. the deal information culture, the and of brain, well the source may primary This the information. into be energy free convert that parameters. gener- thermodynamical the to tying information agency, of in ation produced information bound thermodynamical the a on yields model the particular, In Sec- is paper the for of ways core different The in tion agencies. and of indirectly kinds different realised is this ever, ia hsc ls(h o-eeiiyasmto f the [ of) hypothesis assumption non-genericity statis- past (the of assumption plus genericity physics the tical here mean understood; I the which widely by by the thermodynamics, for accounted of of is principle time-asymmetry turn, second manifest in the This, to world. it macroscopic trace to is tive i)creain ecie ytepyia laws. physical the by probabilis- links, described or physical correlations (deterministic The ignoring tic) physical the in events. of these grounded some of world: is namely description the agency of in of manner events notion a to to solely refers refer it not does agency that h oe rsne eedrvsfo structural a from derives here presented model The oilsrt hsie,cnie h s faec in agency of use the consider idea, this illustrate To eoyadaec a hsb iwda mechanisms as viewed be thus can agency and Memory h e oadestentr faec,i orecognise to is agency, of nature the address to key The I H GN SAPYIA SYSTEM PHYSICAL A AS AGENT THE II. ingnrtdb gnyi h rwhof growth the in agency by generated tion IV .Aec sdseadn hscllinks physical disregarding is Agency A. hr hwhwtedt ewe thermody- between dots the how show I where , inain ii h oreo h infor- the of source the (iii) rientation; oin faec,adosrethat observe and agency, of notions t stebekn fa approximation an of breaking the is y frslwetoyit information. into entropy low sforms tc tutrlsmlrt between similarity structural a otice on fpyis i htmksa makes what (i) physics: of point m smer othermodynami- to asymmetry ime omto nwihbooybuilds. biology which on formation n nai,Cnd,NA5B7. N6A Canada, Ontario, on, tro aaa 2 2Y5. N2L Canada, ntario, 38 asil,France. Marseille, 13288 , 14 ,nml h atta nrp was entropy that fact the namely ], 15 , 16 .A esalse how- see, shall we As ]. 17 ]. 2 them describing the response of physical systems (ther- the physical links are ignored. The agent is the system modynamical or quantum) when they are acted upon in whose physical links are neglected in a given account. certain manners. For instance: ‘If we compress the vol- To see how this works in general, however, we have to ume of a gas, the temperature increases so and so’. Or: distinguish variants of the notion of agency. ‘If we prepare a q-bit in this state, and then measure this spin, we obtain this number’. These are descriptions of behaviours of a part of the world, when an agent acts B. Different notions of agency on it one way or the other. The language of agency is explicit in numerous presentations of these theories, and In a wide sense, any physical system acting on a second is sometimes deemed essential. system and affecting it can be called an agent. But the A moment of reflection, however, shows that this lan- word ‘agency’ is commonly employed in a more restricted guage can be translated away. Any occurrence of ‘if an sense, indicating the capacity of certain systems, such as agent acts on the system in this and that manner’ can humans, to take independent, autonomous, intentional be translated into a statement of the form ‘if the system decisions and act on these. happens to interact in this and that manner’; thus trad- The ambiguity in the use of the term is reflected in the ing the independence of the agent with the modality that philosophical debate about agency (see for instance [18] is at the basis of all physical laws. and references there). From the perspective of physics, Physical laws, indeed, refer to regularities, namely to the ambiguity refers to the assumptions about the rea- repetitive behaviour happening under repeated circum- sons for an agent to act in one way or the other. There is stances. They are generically of the form ‘Anytime that a spectrum of (overlapping) possibilities, leading to dis- A then also B’, or ‘anytime that A then the probability of tinct notions of agency, which can be denoted as follows. B is so and so’. The ‘anytime’ is a conditional (‘if’). The We can call: phase space of classical mechanics and the Hilbert space External agent: any system, when we simply disregard of quantum theory are spaces of possibilities, where the the reasons for its behaviour. Example: in dealing conditionals reside. Laws have been found, in principle, with the dynamics of the Moon’s surface, a me- by generalisation and induction out of a number of re- teorite that impacts on its surface is an external peated observations. Hence the notion of an agent ‘free agent. to act’ is actually irrelevant in the foundations of ther- Internal agent: a system governed by some complex in- modynamics and quantum theory: it can be replaced by ternal dynamics which we could reconstruct. Ex- the conditional: ‘whenever this, then that’. ample: This computer is the agent that controls But the opposite is equally true. And it is more inter- that door. esting. Precisely because physics is modal in this manner, Random agent: a system governed by a genuinely prob- we can always replace the conditionals with the action of abilistic dynamics. an independent agent. And express the arbitrariness by attributing it to something that we call ‘agency’. There- Independent agent: a system governed by an internal fore the agent is here simply the place where we arbitrar- (deterministic or probabilistic) dynamics, too com- ily decide to start the sequence of correlations described plex for us to reconstruct. Example: This man is by the laws we are interested in: it is, in other words, the agent that decides whether to open that door. where we ignore previous physical links. Supernatural agent: a system that does not satisfy nei- To illustrate this, consider for instance the statement ther deterministic nor probabilistic physical laws. that the temperature of a mass of real gas increases when External agency is only a way of talking about external compressed. The compression is due to the interaction physical links when we are not interested in accounting between the gas and some other physical system. This for them. Random agency can be instantiated by quan- other physical system can be a human agent freely decid- tum theory. Human agency is an example of independent ing to push a piston; but also the wind pushing a mass agency [19, 20]. The existence of agency that does not of atmospheric air downhill along a mountain. For the to satisfy neither deterministic nor probabilistic physical gas, which is what is being considered, the difference is laws (supernatural) would contradict our current under- irrelevant: the human and the wind are ‘agents’. What standing of our world I see no interest in considering it. makes them agents, here, is simply the fact that in de- The most interesting case is independent agency, in par- scribing the behaviour of the gas we are not interested in ticular when the agent can choose between alternatives the chain of physical links they might happen to follow: that affect the world differently. In Section IV we shall these are treated as external, arbitrary. It is this that see how a physical system can actually do so. makes them agents here: ignoring their physics. This is in fact general. Agency is always associated to the boundaries of an incomplete or approximate de- III. TIME ORIENTATION scription of the world, within which physical links are to some extent closed, namely approximately sufficient to Agency is time oriented: it affects the future, not the account for the evolution. It refers to the spots where past. What is the source of this time asymmetry? The 3 answer is delicate, because it differs for different notions B. Physical time orientation of agency [15]. It is not difficult to find the source of time-oriented phenomena: the entire macroscopic world around us is manifestly time oriented. We understand this time ori- A. Perspectival time orientation entation of the macro-world in terms of the second prin- ciple (in a generalised sense, and including the past hy- pothesis) which is the only ‘fundamental’ law that breaks Let’s start with the simplest case. Consider an elastic time-reversal invariance. There is no reason for agency collision between a ball B1 and a ball B2. When it is hit to be different, and there is no other source of time ori- by B1, the ball B2 changes its velocity. Say the velocity entation available in our universe (see below for a discus- before the collision was ~vpast and after the collision it sion about quantum theory). Agency must therefore be becomes ~vfuture. We can say that B1 has acted on B2 and a macroscopic phenomenon governed by an entropy gra- the effect of this action is in the future: it has changed dient (and ultimately the past hypothesis of a primordial ~vpast into ~vfuture. This is a possible example of an action low entropy that underpins it) [15, 16]. This must be the affecting the future. ground for the orientation of complex agents like us. However, the physical laws governing the collision are This is the only possible answer to the question of the time reversible. There is nothing in the process itself origin of the time orientation of agency, in the context of that picks up a time direction. At given past ~vpast , it is a naturalistic perspective. The main question I address the future to be affected by the act; but at given future in this paper is how an entropy gradient can give rise to ~vfuture, it is the past to be affected. That is: at fixed the behaviour we recognise as agency. As we shall see past, the world with the collision and the world without in the next Section, the additional ingredients needed for the collision have a different future; while at fixed future, this are surprisingly meagre. the world with the collision and the world without the The subtle point is the fact that it is the macroscopic collision have a different past. We could equally describe world to be time oriented. The micro-history of reality the same history backward in time, with the same laws, happens to be such that in a direction of time (the ‘past’) and say that the effect of the interaction has been to the microstate belonged to a low entropy macrostate. change the velocity from ~vfuture to ~vpast. (To even state time orientation we need to have a notion The reason we say that the collision affects the trajec- of macrostates, namely a coarse graining.) Accordingly, tory of the particle B2 after the collision is only to be agency must be accounted form in terms of a macro- found in the regard we give to the phenomenon. It is we scopic/microscopic distinction, in the sense of statistical who are time oriented. In turn, the reason we take the mechanics and thermodynamics. past as fixed is that we can remember it and we cannot This is not a distinction on the basis of size, scale, or influence it, while we cannot remember the future and number of degrees of freedom; it is a distinction relative we can influence it. Hence we consider the past states of to a set variables (called macroscopic), to which we have the two balls as given, and we say that the effect of the access and that have a partially closed dynamics within collision is in the future. The distinction refers to what some approximation. That is, their behaviour can be we know, not to anything in the phenomenon itself. The approximately determined without involving other vari- distinction is perspectival. As far the phenomenon alone ables. (Here I take low initial entropy, or the past hy- is concerned, it is purely linguistic: we simply call effect pothesis, as given: I do not discuss the possibility for what happens after the collision [21]. itself to be perspectival, which is discussed in [22].) Now, consider an internal agent. If we described it It is tempting to jump from this to saying that this is in complete mechanical terms, the time orientation of all there is to say about the time orientation of agency: agency would again be just a linguistic choice. But, as it is perspectival, agency looks time oriented, but it is we have seen in the previous Section, an agent is precisely only because we see it so. But that would be a mistake. a system of which we are disregarding part the dynamics. The reason is that we have simply displaced the prob- When we describe a human being as an agent, we are ob- lem: the collision does not distinguish cause froms effect, viously not describing its complete microphysics. Hence, but we do. And our distinction is rooted in our own the separation between manifest (macroscopic) degrees of agency, which can affect the future but not the past. The freedom and underlying (microscopic) ones that are not phenomena determined by us and our agency —and with accounted for is constitutive to the notion of independent us a large class of other systems we call internal agents— agency [19]. It is precisely this separation that underpins are definitely not time symmetric. the thermodynamical roots of agency’s time orientation. In particular, to have a different effect on the ball B2, The general situation is therefore clear: the root of a different motion of the ball B1 is needed, while I can time orientation in an independent agent is thermo- now choose between different macroscopic futures given dynamical irreversibility. This underpins independent the same macroscopic past. What is the source of time agency, hence our own sense of openness of the future. orientation in this case? This, in turn, gives us the perspective to read even trivial 4 symmetric interaction in a time oriented manner. asymmetry of the macrophysics, namely the second prin- What is missing is to unravel a mechanism showing ciple. How does this connection work? how the thermodynamical irreversibility can account for Since choosing is irreversible it cannot happen without the time orientation of agency and the openness of the entropy increase. Therefore during the interval ta − tb future it implies. This is what is done in the next section. there must be an entropy increase ∆S > 0. On the other hand, suppose we observe the macroscopic evolution. Be- fore the time ta we have no information about which IV. MODELLING THE THERMODYNAMICAL branch will the system follow. After the time tb we can IRREVERSIBILITY OF AGENCY see which branch has been realised, hence we have novel information. Where does the information come from? Consider an independent agent: a complex unpre- The only possible answer is that I is paid for by the in- dictable macroscopic system. Say that in the interval crease in entropy ∆S. A model illustrating how this can happen was devel- between the times ta and tb it acts on the macroscopic world causing an effect. Say it can choose between N oped in [17] to account for the relation between memory alternatives in its action. Consider the time evolution of and entropy. Let us adapt it here to the present case. the macroscopic state of the world, including the agent Consider two systems: a system A (Agent) at temper- ature T and a system W (World) at a lower temperature itself, and call it Qi(t) with i = 1, ..., N labelling the N a possible evolutions (or ‘branches’) of the macro-world. Tw Qi(t) = Qj (t), for all i, j and t What is puzzling, on the other hand, is why the Ei(t) = E, for t T C∆T ables. We are thus in the case of an internal agent, for N < 2 τa Tw . (10) which it is possible to account for the choice: there is no entropy production in the choice, and the choice is This equation bounds the possibility of choosing between fully determined by the macrophysics. A computer play- alternatives, at given thermodynamical parameters. In ing chess, for instance, choses a move on the basis of particular, it shows that a non-vanishing temperature dif- rules. This is an unproblematic case of causal closure of ference ∆T is needed to have a choice. a macroscopic description. To get a sense of this bound, consider it in a very simple If, on the contrary, high level concepts are not suffi- case. Consider a minimal choice between 2 alternatives, cient to account for behaviour, then we are back to the namely N = 2; using T ≪ τa we have micro/macro context. Something else is doing the choice: if it is not the macrophysics, it must be the microphysics. C∆T ≫ k T (11) w There are always very many micro-histories compatible where we have reinserted the Boltzmann constant k =16 with any given high-level account, leaving space for the for clarity. The left hand side of this equation is the branching. excess thermal energy in the agent, while the right hand Neither case conflicts with the causal closure of the mi- side is the average energy per degree of freedom in the croscopic physics. Ultimately, agency is always nothing world. That is: in order to be able to choose, the agents else than ignoring some physical links. must have enough energy to stand up above the thermal energy of the world. Crucially, there is no reduction of entropy in choosing: B. Quantum theory there is increase in entropy, contrary to what appears in the picture where the physics of the agent is disre- I have framed the discussion in terms of classical me- garded. Choosing is a conventional irreversible process, chanics, because including probabilities complicates the and it happens because it is statistically favoured, as all language. But nothing substantial changes in the above irreversible processes do. if quantum mechanics is taken into account. Before concluding, we comment on two points that we Microscopic time reversal invariance is not broken by left open: top-down causation and the role of quantum quantum randomness [23, 24]. The predictions of quan- theory. tum mechanics are formulated in terms transition proba- bilities. These do not distinguish between past and future and are time reversal invariant (CPT invariant in quan- A. Top down causation tum field theory). The discussion in this paper, on the other hand, clarifies the origin of the time asymmetry in Defenders of top-down causation point out that it may our conventional use of quantum theory. We routinely be possible to account for the choice of an agent in terms interpret quantum transition probabilities as time ori- of high-level concepts. For instance, a choice can be mo- ented, namely we routinely read them as probabilities tivated by a value system, a calculation about the future, for future events given past events; but this is perspec- 6 tival. It is because we are agents that can influence the agency, it is the agent that leaves a trace on the exter- future, immersed in a time oriented macroscopic world, nal macroscopic world; while memory is a trace left by that we do so. Therefore the time orientation of the com- the world on the memory system. Both phenomena need mon reading of quantum probabilities is just perspectival. long thermalization times, namely quasi-stable system, As shown, this perspectival time orientation, in turn, is to hold the memory or the effect of the action. Both need ultimately sourced by the arrow of time of the second a disequilibrium in the past, to account for orientation principle, via our own agency. and irreversibility. Both can be understood as macro- Quantum theory does not change anything regarding scopic phenomena pertaining to a coarse grained picture the distinction between microphysics and macrophysics, of the world, and make no sense at the microscopic level either. For the sake of the current discussion quantum (except in metaphorical ‘anthropocentric’ language). indeterminism can be treated as due to unaccounted de- Agency is time oriented because it is a macroscopic grees of freedom. If one wish to, one can even do so ex- phenomenon driven by an entropy gradient (hence ulti- plicitly by using an interpretation of quantum theory like mately by the past hypothesis). The model presented in the de Broglie-Bohm hidden variable one, where indeter- Section IV shows that system separation, past tempera- minism is indeed statistical ignorance, or Many Worlds, ture difference, and long thermalization times are meagre where indeterminism is indexical, namely ignorance of elements nevertheless sufficient to model this thermody- the branch in which we are located. Alternatively, one namical roots of agency. In turn, our time orientation as may simply remember that in order to affect the macro- agents compels us to look at mechanical interactions in world, quantum indeterminism needs decoherence, which a time oriented manner [15]. is precisely based on disregarding degrees of freedom. The most interesting aspects of the two phenomena is Whether the causal closure of the macroscopic descrip- that they both produce information. In agency, infor- tion of the world is in principle accounted by some under- mation can be recognised, in Shannon’s sense, as the in- lying classical deterministic microphysics or by quantum stantiation of one among a number of possibilities. In the randomness is irrelevant for the understanding of agency. case of memory, information is what Shannon calls ‘rel- ative information’: physical correlation between a past macroscopic event and its trace. V. CONCLUSIONS: MEMORY AND THE In both cases, the information is generated by increas- CREATION OF INFORMATION ing entropy. Eq (10) gives the maximal information that can be produced in choosing, at given thermodynamical We have a strong feeling that we cannot influence the parameters. It analogous to bound on the information past, but we can influence the future. This seems to con- produced by the formation of memory derived in [17]. flict with the time (CPT) reversal invariance of funda- Low entropy is a form of information because a lower mental physics. But is not. We have this feeling because entropy state amounts to a more selective information truly we can affect the macroscopic future but not the about the microphysics (a zero entropy macrostate is macroscopic past. The macroscopic world we work with a state that has maximal information about the mi- has a fixed past determined by abundant present traces crophysics: the microstate is unique). Memory and and memories [17], while it is compatible with a number agency utilise the information stored in low entropy and of different futures, that do depend on what happens in translate it into information readable in the macroscopic our brain. This is the openness of the future that our world. In fact, they both can be viewed as mechanisms feeling veridically captures. that generate macroscopic information. This openness of the future leaves ample space for sub- Macroscopic information, stored in human memory, in tle high level processes to influence the macroscopic fu- DNA molecules, in computer messages, in books, in nar- ture. Our sense of being free to decide is clearly rooted ratives, in software codes, in records of any form, must here. It is in this sense, that, as Ismael puts it: ‘Physics have been ultimately produced by physical mechanisms. Makes Us Free’ [20]. Traces of the past and decisions by agents —possibly in The microscopic account is a wholly different story, but turn themselves affected by memories of the past— are is of little relevance for our experience and feelings, since, major sources of everything we call information. In both by definition, we do not access it. cases, information is created, in a statistically favoured Independent agency is a description of the macroscopic manner, at the expenses of low entropy, in accordance dynamics of an interaction between an agent and the with the second principle. In a fully thermalised situa- world which: (i) is unpredictable, (ii) is irreversible, tion, there is no space for memories or for agents. (iii) produces a (macroscopically) detectable effect on the The entire informational universe formed by the world in the future, and (iv) produces information. There biosphere and by culture can therefore perhaps be are remarkable similarities between this and the model viewed, from this perspective, as formed by information for traces, or ‘memories’, described in [17]. produced by a mechanism of the form described here. Both memory and agency are events that leave a trace in the macroscopic domain. The difference is that the *** roles are in a certain sense exchanged: in the case of I am deeply indebted with Jenann Ismael: several 7 ideas of this paper developed in conversations with her. FQXi-RFP-1818 and of the ID# 61466 grant from the Thanks to David Albert for a very insightful criticisms to John Templeton Foundation, as part of the The Quan- an early draft of this paper, crucial for me. This work was tum Information Structure of Spacetime (QISS) Project made possible through the support of the FQXi Grant (qiss.fr). [1] J. Maxwell, Theory of Heat. Longmans, Green, and Co., London, 1871. [2] W. C. Myrvold, “The Science of θ∆cs,”. [3] C. A. Fuchs and B. C. Stacey, “QBism: Quantum theory as a hero’s handbook,” in Proceedings of the International School of Physics ‘Enrico Fermi’, vol. 197, pp. 133–202. dec, 2019. arXiv:1612.07308. [4] F. Ramsey, “General Propositions and Causality,” in Foundations: Essays in Philosophy, Logic, Mathematicvs and Economics, D. Mellor, ed., pp. 133–151. Routledge anf Kegan, London, 1978. [5] H. Price, “Causal Perspectivalism,” in Causation, Physics, and the Constitution of Reality: Russell’s Republic Revisited, pp. 250–292. 2007. [6] J. Pearl, Causality: Models, reasoning and Inference. Cambridge University Press, 2000. [7] J. Woodward, “Causation and Manipulability https://plato.stanford.edu/archives/win2016/entri,” 2016. https://plato.stanford.edu/archives/win2016/entries/causation-mani/. [8] D. Deutsch, “Constructor theory,”Synthese 190 (dec, 2013) 4331–4359, arXiv:1210.7439. [9] M. Kirchhoff, T. Parr, E. Palacios, K. Friston, and J. Kiverstein, “The Markov blankets of life: autonomy, active inference and the free energy principle,” J R Soc Interface 15 (2018) 20170792. [10] A. Kolchinsky and D. Wolpert, “Semantic information, autonomous agency and non-equilibrium statistical physics,” Focus 8 (2018) 20180041. [11] S. Still, “Thermodynamic Cost and Benefit of Memory,” Physical Review Letters 124 (2020) 050601. [12] I. Durham, “A Formal Model for Adaptive Free Choice in Complex Systems,” Entropy 22 (2020) 568. [13] B. Loewer, “The Consequence Argument Meets the Mentaculus,”. http://philsci-archive.pitt.edu/17328/. [14] D. Albert, Time and Change. Harvard University Press, Cambridge MA, 2000. [15] H. Price, Time’s Arrow. Oxford University Press, 1996. [16] D. Albert, “The Difference between the Past and the Future,” in After Physics. Harvard University Press, 2015. [17] C. Rovelli, “Memory and entropy,” arXiv:2003.06687. http://arxiv.org/abs/2003.06687. [18] M. Schlosser, “Agency,” 2019. https://plato.stanford.edu/archives/win2019/entries/agency/. [19] B. Spinoza, Ethics, Third Part, Proposition II and Scolio. 1677. http://en.wikisource.org/wiki/Ethics_(Spinoza)/Part_3. [20] J. Ismael, How Physics Makes Us Free. Oxford University Press, 2016. [21] D. Hume, Treatise of Human Nature. 1736. [22] C. Rovelli, “Is Time’s Arrow Perspectival?,” in The Philosophy of Cosmology, K. Chamcham, J. Silk, J. Barrow, and S. Saunders, eds., pp. 285–296. Cambridge University Press, 2016. arXiv:1505.01125. [23] A. Einstein, R. Tolman, and B. Podolsky, “Knowledge of past and future in quantum mechanics,” Phys. Rev. 37 (1931) 780. [24] C. Rovelli, “An Argument Against the Realistic Interpretation of the Wave Function,” Foundations of Physics 46 (2016) no. 10, 1229–1237, arXiv:1508.05533.
Arxiv:2007.05300V2 [Physics.Hist-Ph] 13 Jul 2020
![Arxiv:2007.05300V2 [Physics.Hist-Ph] 13 Jul 2020](http://data.docslib.org/img/5a44d7e4d69c6976ef7ee63adb061cbc-1.webp)