Are There Experimental Arguments Independent of Theories? In Defense of a Hackingian Approach to the Scientific Realism Debate

ABSTRACT

This paper defends a Hackingian approach to the scientific realism debate by arguing against mainstream realists’ and antirealists’ common claim that no experimental arguments for the reality of posited entities can be theory-independent. Opposing this claim, I argue that some experimental arguments can warrant belief in the existence of entities without depending on the truth of the theories that posit the entities and describe their properties and the theories that explain the interactions between the entities and the experimental devices. To achieve this goal, I investigate Hacking’s case of electron spraying, the famous experiment of positioning atoms by means of a scanning tunnelling microscope, and some novel experiments that use Ashkin’s optical tweezers. By introducing the distinction between theoretical properties and ontic properties, I argue that we can build up some theory-independent criteria of reality on the basis of ontic properties. Manipulability is an ontic property and the manipulation criterion is thus a theory-independent one, which should be reexamined and used for those experimental cases along a local approach to the realism debate. Resisting the mainstream realists’ and antirealists’ objections, I contend that some experiments using optical tweezers can warrant belief in the existence of the experimented entities without depending on theories.

Acknowledgements

I am deeply grateful to two anonymous reviewers for their very valuable comments that substantively helped improve this article. I also thank Jonathon Hricko for his very helpful comments and suggestions on the draft and on the revised version. Finally, I would like to express my greatest gratitude toward Professor Ian Hacking who passed away on 10 May 2023. Professor Hacking visited Taiwan in 2007. During the period of his staying in Taiwan, I had many interactions with him and was strongly impacted by his view, which led me to the scientific realism debate via the experimental approach. This article is dedicated to commemorate him.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Miller (2016, 994) points out that Hacking’s home truths ‘consist of thinly theory-laden low-level generalizations about [experimental entities’] well understood causal properties; for example, their mass, spin, and charge.’ However, the properties of electrons such as mass, spin, and charge are just theoretical. Hence, Hacking’s case of electron-spraying must involve theories about electrons.

2 According to the same reason, the manipulation criterion cannot be appropriately applied to Gelfert’s case of quasi-particles. ‘Quasi-particles’, by definition, refers to the collective excitations of a many-particles system (for example, an ensemble of electrons). ‘These excitations are indistinguishable in terms of their manipulability from the actual particles.’ (Gelfert 2003, 257) If one needs theories to interpret the manipulation of actual particles (for example, electrons) and their ontological status, then one necessarily needs theories to interpret the manipulation of quasi-particles and their ontological status. Both depend on theories in the sense of T1 or T2.

3 Chakravartty also connects the notion of detection properties to the notion of structure in structural realism. I put this point aside. For my purposes, the important point is that the notion of detection properties is theoretical.

4 See ‘Groundbreaking inventions in laser physics,’ from the webpage, Scientific Background on the Nobel Prize in Physics 2018, https://www.nobelprize.org/uploads/2018/10/advanced-physicsprize2018.pdf.

5 In the experiment published in 1987, Ashkin reports ‘the experimental demonstration of optical trapping and manipulation of individual viruses and bacteria in aqueous solution by laser light using single-beam gradient force traps’ (Ashkin and Dziedzic 1987, 1517). This experiment was conducted to explore the potentiality of the optical trapping instrument. The question that Ashkin wanted to answer is whether or not the instrument can trap living organisms without damaging them.

6 There is always a veridicality issue regarding microscopic images when using microscopic observation to demonstrate the reality of unobservable entities. Hacking (1985, 146–148) proposes the argument of the grid to endorse the validity of microscopic images as a demonstration of reality. He argues that ‘I know that what I see through the microscope is veridical because we made the grid to be just that way’ (148). van Fraassen (1985, 298) argues that Hacking’s argument commits to a circularity, because the premise implies that we successfully made the object to be that way. This dispute does not endanger the use of microscopic observation in my case, because I argue that scientists demonstrate the existence of entities by observing the manipulation rather than by observing the entities and their properties. That is, scientists offer two series of images of the DNA molecule at two different time intervals. The two series contain many different images at different times, by which we observe the manipulation process of the DNA molecule rather than its shape or structure. In fact, the real shape and structure of the DNA are still unobservable. Given the two series of images, one has a good reason to accept that the DNA has been manipulated even if one cannot demonstrate that each individual static image, considered in isolation from the other images, is veridical. Of course, van Fraassen does not accept that observation through microscopes can provide reliable access to the reality of entities that are unobservable by unaided eyes. However, other anti-realist such as Bueno don’t agree with van Fraassen’s extreme position, because he admits that ‘[m]icroscopes’ images are often taken as evidence for the existence of certain unobservable entities’ (Bueno 2011, 424). A sufficient discussion about this issue needs another occasion.

7 One may still be skeptical about the theory-independence of observation through microscopes. Indeed, the explanation of observation through microscopes depends on the working principles of optical instruments. However, these working principles are neither the theories positing the manipulated entities nor the theories explaining how optical tweezers interact with the entities. In those experiments, the microscopes were arranged outside of the main experimental apparatus. Microscopes were not used to manipulate the experimented objects and only passively received the light reflected from them. We do not need the working principles of optical instruments to judge that some things have been manipulated because scientists really see some fixed images or changes of images through their microscopes.

Additional information

Funding

This work was supported by Ministry of Science and Technology, Taiwan: [grant number MOST108-2410-H-194-002-MY3].