Organicism and Reductionism in Cancer Research: Towards a Systemic Approach

Abstract

In recent cancer research, strong and apparently conflicting epistemological stances have been advocated by different research teams in a mist of an ever‐growing body of knowledge ignited by ever‐more perplexing and non‐conclusive experimental facts: in the past few years, an ‘organicist’ approach investigating cancer development at the tissue level has challenged the established and so‐called ‘reductionist’ approach focusing on disentangling the genetic and molecular circuitry of carcinogenesis. This article reviews the ways in which ‘organicism’ and ‘reductionism’ are used and opposed in this context, with an aim at clarifying the debate. Methodological, epistemological and ontological implications of both approaches are discussed. We argue that the ‘organicist/reductionist’ opposition in the present case of carcinogenesis is more a matter of diverging heuristics than a claim about theoretical or ontological (ir)reducibility. As a matter of fact, except for the downward causation claim, which we question, we argue that the organicist arguments are compatible with the reductionist approach. Moreover, we speculate that both approaches, which currently focus on specific entities i.e., genes versus tissues, will need to shift their conceptual frameworks to studying complex arrays of relationships potentially ranging over several levels of entities, as is the case with ‘systems biology’.

Acknowledgements

This paper was triggered by extremely fruitful discussions with Ana Soto and Carlos Sonnenschein following seminars and meetings keenly organized by Paul‐Antoine Miquel. I also thank Michel Morange and Pierre‐Alain Braillard for insightful comments on an earlier draft, and Jean Gayon for his thoughtful mentoring.

Notes

[1] The International Agency for Research on Cancer, which is part of the UN World Health Organization, estimates the European mortality rate due to cancer at some 1.7 million deaths per year (Boyle and Ferlay 2005); other estimates point to 10 million deaths per year worldwide (Khayat 2005).

[2] Cancer research is an extremely complex field of scientific investigation. In this paper, we focus on two ways of looking at cancer development as illustrated by two specific research teams. This bipolarization should not be seen as an oversimplification of the cancer research field nor of its history: it serves as an anchor point only for our philosophical discussion on organicism and reductionism. For historical views on cancer research, see for instance van Helvoort (1999) or Morange (1997).

[3] A weaker formulation of this reduction would focus solely on single explanations (and not on all explanations formulated in a particular field of science). In this case, the reduction would occur between explanations (and not between fields of science): if an explanation is shown to be a special case of another explanation, then the first explanation could be said to be reduced to the second one in the weaker sense of ‘explanation reduction’.

[4] To illustrate this possibility, one could imagine the following fictitious scenario:

1. E₁ and E₂ are true explanations of ‘why Paul got cancer’,

2. E₁ appeals to molecules (e.g., molecular expression), E₂ to tissue characteristics (e.g., wound),

3. The contrast class more specifically highlights ‘what was specific about Paul history that made him get cancer’,

4. While it is known that ‘Paul was wounded and later a cancer grew at the very place of the wound’.

In this case, E₂ is a better explanation than E₁ since it fits perfectly well the contrast class (C) and the relevance context (D). However, one could imagine also that:

• (C’) The contrast class more specifically highlights ‘how is it that Paul’s wound led to cancer?’

In this alternative scenario, explanation E₁ would be the better one since it would describe the molecular mechanisms at stake and how, for instance, a wound could impact the molecular expression of certain genes and thereby lead to cancer.

In this fictitious scenario, we see the possibility for the would‐be TOFT explanation E₂ to be a good explanation whereas the SMT explanation E₁ would not be good or the other way around: the difference simply comes from the explanatory contrast class and the relevance context.

[5] The other three versions can be considered as variants of these three major ones and are not of central interest to our discussion: in addition to weak emergence, synchronic emergence and diachronic structure emergence, Stephan identifies weak diachronic emergence, strong diachronic emergence and strong diachronic structure emergence. Please refer to Stephan (1999) for more details.

[6] The example under consideration concerns the work of biologist Emmanuel Farge on how developmental gene expression can be mechanically regulated by morphogenetic movements. In his work, Farge demonstrates how the application of a mechanical force—or a pressure—on certain surface cells of a Drosophila embryo induces the expression of the ‘Twist gene’ around the entire dorsal‐ventral axis and results in the ventralization of the embryo (see Brouzès and Farge 2004; Farge 2003).

[7] In a similar fashion, Rosenberg and Kaplan (2005) argues that the formation or mal‐formation of a chick‐limb can be accounted for simply by appealing to upward causation, whereas Kitcher (1984) had presented this example as a biological case of downward causation. Also, in cancer research, Jacks and Weinberg (2002) provide an ‘upward‐causation compatible’ interpretation of the specific work of Bissell et al. initially formulated as a matter of tissue organization. Morange (2005) also believes that an upward‐causation explanation of the example borrowed by Soto and Sonnenschein is fully possible.

[8] Bedau studies cellular automata and in particular Conway’s ‘game of life’ (Bedau 1997); he calls ‘weak emergence’ the form of diachronic emergence exhibited by such models of complex systems. To avoid confusion with Stephan’s typology of emergence (Stephan 1999), we prefer to use ‘diachronic structure emergence’ in this case.

Additional information

Notes on contributors

Christophe Malaterre

Christophe Malaterre is at the Institut d’histoire et de philosophie des sciences et des techniques (IHPST), Université Paris 1. Panthéon Sorbonne.