Bionic systems, connecting biological tissues with computer or robotic devices through brain–machine interfaces, can be used in various ways to discover biological mechanisms. In this article I outline and discuss a “stimulation-connection” bionics-supported methodology for the study of the brain, and compare it with other epistemic uses of bionic systems described in the literature. This methododology differs from the “synthetic”, simulative method often followed in theoretically driven Artificial Intelligence and cognitive (neuro) science, even though it involves machine models of biological systems. I also bring the previous analysis to bear on some claims on the epistemic value of bionic technologies made in the recent philosophical literature. I believe that the methodological reflections proposed here may contribute to the piecewise understanding of the many ways bionic technologies can be deployed not only to restore lost sensory-motor functions, but also to discover brain mechanisms.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Tax calculation will be finalised during checkout.
The simulation-replacement methodology is called ArB, from “Artificial replacement of Biological components”, in (Datteri 2009). The “simulation-replacement” label is used here to emphasize some of the main differences between it and the “stimulation-connection” method, as discussed later on.
The expression “mechanistic model” is used here to refer to the description of a mechanism (Craver 2007). In what follows I assume that mechanistic models describe the regular behaviour of system components by means of generalizations (Glennan 2005; Woodward 2002). The term “model” is used to emphasize the fact that mechanism descriptions may be more or less abstract in the sense clarified by Suppe (1989): they characterize the behaviour of each component as depending on a restricted (though not necessarily narrow) set of factors. For example, a model might characterize the activity of the neurons in a particular brain area as depending only on the firing rate of neurons in another area; a less abstract model would take into account more input or boundary factors. Both models express counterfactual generalizations stating that the behaviour of reticular neurons would be such and such, if it depended only on that restricted set of factors (Suppe 1989; Woodward 2002). By making these epistemological assumptions I am not claiming that the ensuing methodological discussion of the stimulation-connection methodology is consistent only with this interpretation of the notion of a “mechanistic model”.
To be more precise, they have inhibited the activity of this component by using a particular experimental apparatus. See the cited article itself for further detail.
For an example of such a comparison, see the graphs showed in Fig. 2 of the cited article.
The distinction made here can be brought to bear on the question of what exactly is a bionics-supported neuroscientific experiment. It may be defined as a neuroscientific experiment exploiting in vivo connections between living systems and artificial devices. This definition would be too liberal, however, as traditional neurophysiological experiments—e.g., voltage clamp experiments—would fit it. On the contrary, it would be too restrictive to include in the class of bionics-supported experiments only those experiments in which the target living system controls robotic devices, as this would exclude experiments in which the subject brain-controls virtual devices (as, for example, in the set-up illustrated in Fig. 2). Then, one may include in that class all and only those neuroscientific experiments which involve hybrid systems whose structure can be described as in Fig. 4. It is worth noting, however, that the label “hybrid bionic system” is typically used in the contemporary scientific literature to refer to systems in which the artificial device functionally replaces a biological component. That is to say, contemporary bionics research is focused on the realization of devices which are essential for people suffering from motor or sensory limitations to perform particular tasks, as they can replace sensory or motor organs. For this reason, here I will restrict the label “bionics-supported neuroscientific experiment” to all and only the experiments which make an essential use of artificial devices qua replacers of biological components. The label therefore does not apply to the experiments carried out in the “pole control” phase of the monkey study, even though they have involved a system that could be structurally described as in Fig. 4. As suggested by the analysis made in this article, overlooking this distinction may obscure the methodological novelty of the experiments carried out in the “brain control” phase of the monkey study.
To be sure, experiments in electrophysiology often involve artificial stimulations of the target biological tissue. For example, one may intervene on the membrane potential of particular neurons after blocking specific kinds of ion channels in order to find the threshold above which action potentials are generated in those conditions. Note that, in experiments of this sort, the nature and magnitude of the “input” stimulation (e.g., of changes in membrane potential) do not systematically depend on the effects of that stimulation (e.g., on whether action potentials are generated or not). The “input” parameters are independent of the “output” of the interventions: the experimenter explores a relevant portion of the “input” space and measure the effects in order to find a correlation. Quite on the contrary, in non-simulation, connection bionic experiments, the nature and magnitude of the stimulation received by the biological system crucially depends on the “output” of biological activity—that is to say, on the behaviour of the prosthesis as determined by the biological system itself. Plastic changes in the subject’s brain depend on the feedback informing the subject about the way her brain is moving the prosthesis. Brain activity determines prosthetic behaviour; information on prosthetic behaviour determines changes in brain activity. Such a “circular” connection between the nature and magnitude of the stimulations applied to brain circuits and the nature and magnitude of the effects of those stimulations is not established in traditional electrophysiological intervention experiments.
Datteri (2009) is explicit in placing the “no-plasticity” constraint on the simulation-replacement method only. This constraint is called “ArB2” there, ArB being the label used to refer to the simulation replacement methodology. It is one of the constraints under which “H may provide experimental support for the hypothesis that component b1 (i.e., the component removed from B to obtain H) behaves as MB prescribes” (p. 310, emphasis added), H, B, and MB being the hybrid system, the target biological system, and the mechanistic model under scrutiny respectively.
Note that the simulation replacement method has, to the best of my knowledge, only been applied once, namely in the lamprey study reported in (Zelenin et al. 2000). At some point in her article, Chirimuuta voices the suspicion that the methodological analysis offered in (Datteri 2009) has nothing to say about a vast class of bionic experiments reported in the literature. While looking outside of the mainstream can point up novel research approaches, supplementing earlier analyses and stimulating methodological discussion of novel and emerging fields previously overlooked by philosophers, I agree with her on the fact that (Datteri 2009) covers a minimal part of the contemporary bionics-supported research. The aim of this article is to examine the structure of a methodology which is much more often adopted in contemporary scientific research than the simulation-replacement one.
Carmena, J. M., Lebedev, M. A., Crist, R. E., O’Doherty, J. E., Santucci, D. M., Dimitrov, D. F., et al. (2003). Learning to control a brain–machine interface for reaching and grasping by primates. PLoS Biology, 1(2), 193–208.
Chapin, J. K., Moxon, K. A., Markowitz, R. S., & Nicolelis, M. A. L. (1999). Real-time control of a robot arm using simultaneously recorded neurons in the motor cortex. Nature Neuroscience, 2(7), 664–670.
Chirimuuta, M. (2013). Extending, changing, and explaining the brain. Biology and Philosophy, 28(4), 613–638.
Chou, P. C., & Hannaford, B. (1997). Study of human forearm posture maintenance with a physiologically based robotic arm and spinal level neural controller. Biological Cybernetics, 76(4), 285–298.
Cordeschi, R. (2002). The discovery of the artificial behavior mind and machines before and beyond cybernetics. Dordrecht: Springer.
Craver, C. (2007). Explaining the brain: Mechanisms and the mosaic unity of neuroscience. Oxford: Clarendon Press.
Craver, C. F. (2010). Prosthetic Models. Philosophy of Science, 77(5), 840–851.
Datteri, E. (2009). Simulation experiments in bionics: a regulative methodological perspective. Biology and Philosophy, 24(3), 301–324.
Datteri, E., & Tamburrini, G. (2007). Biorobotic Experiments for the Discovery of Biological Mechanisms. Philosophy of Science, 74(3), 409–430.
Floreano, D., Ijspeert, A. J., & Schaal, S. (2014). Robotics and Neuroscience. Current Biology, 24(18), R910–R920.
Ganguly, K., & Carmena, J. M. (2009). Emergence of a stable cortical map for neuroprosthetic control. PLoS Biology, 7(7), e1000153.
Ganguly, K., Dimitrov, D. F., Wallis, J. D., & Carmena, J. M. (2011). Reversible large-scale modification of cortical networks during neuroprosthetic control. Nature Neuroscience, 14(5), 662–667.
Glennan, S. (2005). Modeling mechanisms. Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences, 36(2), 443–464.
Golub, M. D., Chase, S. M., Batista, A. P., & Yu, B. M. (2016). Brain–computer interfaces for dissecting cognitive processes underlying sensorimotor control. Current Opinion in Neurobiology, 37, 53–58.
Grasso, F. W., Consi, T. R., Mountain, D. C., & Atema, J. (2000). Biomimetic robot lobster performs chemo-orientation in turbulence using a pair of spatially separated sensors: Progress and challenges. Robotics and Autonomous Systems, 30(1–2), 115–131.
Hochberg, L. R., Serruya, M. D., Friehs, G. M., Mukand, J. A., Saleh, M., Caplan, A. H., et al. (2006). Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature, 442(7099), 164–171.
Koralek, A. C., Jin, X., Long, J. D., II, Costa, R. M., & Carmena, J. M. (2012). Corticostriatal plasticity is necessary for learning intentional neuroprosthetic skills. Nature, 483(7389), 331–335.
Le Masson, G., Renaud-Le Masson, S., Debay, D., & Bal, T. (2002). Feedback inhibition controls spike transfer in hybrid thalamic circuits. Nature, 417(6891), 854–858.
Lebedev, M. A., Carmena, J. M., O’Doherty, J. E., Zacksenhouse, M., Henriquez, C. S., Principe, J. C., et al. (2005). Cortical ensemble adaptation to represent velocity of an artificial actuator controlled by a brain–machine interface. The Journal of Neuroscience, 25(19), 4681–4693.
Newell, A., & Simon, H. A. (1961). Computer Simulation of Human Thinking: A theory of problem solving expressed as a computer program permits simulation of thinking processes. Science, 134(3495), 2011–2017.
Nicolelis, M. A. L. (2003). Brain–machine interfaces to restore motor function and probe neural circuits. Nature Reviews Neuroscience, 4(5), 417–422.
Nicolelis, M. (2011). Beyond Boundaries: The New Neuroscience of Connecting Brains With Machines And How It Will Change Our Lives. New York: Times Books.
Orsborn, A. L., & Carmena, J. M. (2013). Creating new functional circuits for action via brain–machine interfaces. Frontiers in Computational Neuroscience, 7, 157.
Rosenblueth, A., & Wiener, N. (1945). The Role of Models in Science. Philosophy of Science, 12(4), 316–321.
Simon, H. A., & Newell, A. (1962). Computer Simulation of Human Thinking and Problem Solving. Monographs of the Society for Research in Child Development, 27(2), 137–150.
Suppe, F. (1989). The Semantic Conception of Theories and Scientific Realism. Urbana and Chicago: University of Illinois Press.
Tamburrini, G., & Datteri, E. (2005). Machine Experiments and Theoretical Modelling: from Cybernetic Methodology to Neuro-Robotics. Minds and Machines, 15(3–4), 335–358.
Wander, J. D., & Rao, R. P. N. (2014). Brain–computer interfaces: A powerful tool for scientific inquiry. Current Opinion in Neurobiology, 25, 70–75.
Webb, B. (2001). Can robots make good models of biological behaviour? The Behavioral and Brain Sciences, 24(6), 1033–1050.
Wessberg, J., Stambaugh, C. R., Kralik, J. D., Beck, P. D., Laubach, M., Chapin, J. K., et al. (2000). Real-time prediction of hand trajectory by ensembles of cortical neurons in primates. Nature, 408(6810), 361–365.
Woodward, J. (2002). What Is a Mechanism? A Counterfactual Account. Philosophy of Science, 69, S366–S377.
Zelenin, P. V., Deliagina, T. G., Grillner, S., & Orlovsky, G. N. (2000). Postural control in the lamprey: A study with a neuro-mechanical model. Journal of Neurophysiology, 84(6), 2880–2887.
About this article
Cite this article
Datteri, E. The Epistemic Value of Brain–Machine Systems for the Study of the Brain. Minds & Machines 27, 287–313 (2017). https://doi.org/10.1007/s11023-016-9406-1
- Brain–machine interfaces
- Prosthetic models
- Bionic experiments in neuroscience
- Robot-based simulation methodologies
- Discovering mechanisms in neuroscience
- Artificial Intelligence