Let a basic physical property be any simple, unrealized property that features in the laws of ideal completed physics.1 Perhaps electric charge and mass-energy are such properties.Proponents of powers ontologies 2 think that at least some basic physical properties have at least partially causal or dispositional essences. Powers ontologies are motivated by problems with categoricalism, the view that basic physical properties have no non-trivial essential modal character. There are two main categoricalist positions on the relationship between properties and laws: the Mill-Ramsey-Lewis best systems theory (MRL), and the Dretske-TooleyArmstrong nomic necessitation theory (DTA), which can be characterised as reductionist and anti-reductionist, respectively. 3 According to MRL, laws of nature reduce by supervenience to local matters of contingent fact. Roughly, the laws of nature are the theorems of the deductive system with the best possible combination of simplicity and strength in relation to the totality of available evidence. 4 On Lewis' version of MRL, basic categorical properties derive their causal roles from the total pattern of instantiation of such properties throughout spacetime, upon which pattern the laws of nature supervene. On DTA, second-order lawmaking relations are introduced to explain why first-order universals have their causal profiles-Fs cause Gs in virtue of N(F,G). Laws are irreducible second-order governing relations, and confer causal roles upon intrinsically inert categorical properties. 1 The reason I use 'basic' rather than 'fundamental' will become clear in §3. I also use 'basic' to describe (putatively) non-composite physical particulars such as electrons and quarks. I likewise apply 'complex' to both properties and particulars.2 Shoemaker (1980); Ellis (2002);Heil (2003); Molnar (2003); Mumford (2004); Bird (2007a); Martin (2007). 3 For a canonical statement and defence of MRL, see Lewis (1994); for DTA see Armstrong (1983). 4 Lewis (1994), pp. 478-9. forthcoming in Synthese (Penultimate draft. See published version at http://dx.doi.org/10.1007/s11229-017-1417-9) 2 There are many well-rehearsed problems with these positions. Given that MRL laws supervene on local matters of fact, how can they also explain those matters of fact? How can the laws play a role in bringing about the particular events they subsume, if the laws themselves are metaphysically grounded in the total course of events? Is what counts as the best system in part dependent on the quirks of human psychology, and if so, is MRL inconsistent with the objectivity required by robust scientific realism? 5 If the N-relation in DTA's N(F,G) doesn't essentially ground the fact that Fs cause Gs, then don't we need a third-order N-relation to explain why it does at our world, which sets us off on a regress? Conversely, if we say that N(F,G) is essential to N, then N itself can't be categorical. 6 Finally, most versions of categoricalism, assuming in addition that categorical properties stand in primitive transworld ident...