Oxford University Press's
Academic Insights for the Thinking World

Why causality now?

Head hits cause brain damage, but not always. Should we ban sport to protect athletes? Exposure to electromagnetic fields is strongly associated with cancer development. Should we ban mobile phones and encourage old-fashioned wired communication? The sciences are getting more and more specialized and it is difficult to judge whether, say, we should trust homeopathy, fund a mission to Mars, or install solar panels on our roofs. We are confronted with questions about causality on an everyday basis, as well as in science and in policy.

Causality has been a headache for scholars since ancient times. The oldest extensive writings may have been Aristotle, who made causality a central part of his worldview. Then we jump 2,000 years until causality again became a prominent topic with Hume, who was a skeptic, in the sense that he believed we cannot think of causal relationships as logically necessary, nor can we establish them with certainty.

The next major philosophical figure after Hume was probably David Lewis, who proposed quite a controversial account saying roughly that something was a cause of an effect in this world if, in other nearby possible worlds where that cause didn’t happen, the effect didn’t happen either. Currently, we come to work in computer science originated by Judea Pearl and by Spirtes, Glymour and Scheines and collaborators.

All of this is highly theoretical and formal. Can we reconstruct philosophical theorizing about causality in the sciences in simpler terms than this? Sure we can!

One way is to start from scientific practice. Even though scientists often don’t talk explicitly about causality, it is there. Causality is an integral part of the scientific enterprise. Scientists don’t worry too much about what causality is­ – a chiefly metaphysical question – but are instead concerned with a number of activities that, one way or another, bear on causal notions. These are what we call the five scientific problems of causality:

8529449382_85663d5f6a_o
Phrenology: causality, mirthfulness, and time. Photo by Stuart, CC-BY-NC-ND-2.0 via Flickr.
  • Inference: Does C cause E? To what extent?
  • Explanation: How does C cause or prevent E?
  • Prediction: What can we expect if C does (or does not) occur?
  • Control: What factors should we hold fixed to understand better the relation between C and E? More generally, how do we control the world or an experimental setting?
  • Reasoning: What considerations enter into establishing whether/how/to what extent C causes E?

This does not mean that metaphysical questions cease to be interesting. Quite the contrary! But by engaging with scientific practice, we can work towards a timely and solid philosophy of causality.

The traditional philosophical treatment of causality is to give a single conceptualization, an account of the concept of causality, which may also tell us what causality in the world is, and may then help us understand causal methods and scientific questions.

Our aim, instead, is to focus on the scientific questions, bearing in mind that there are five of them, and build a more pluralist view of causality, enriched by attention to the diversity of scientific practices. We think that many existing approaches to causality, such as mechanism, manipulationism, inferentialism, capacities and processes can be used together, as tiles in a causal mosaic that can be created to help you assess, develop, and criticize a scientific endeavour.

In this spirit we are attempting to develop, in collaboration, complementary ideas of causality as information (Illari) and variation (Russo). The idea is that we can conceptualize in general terms the causal linking or production of effect by the cause as the transmission of information between cause and effect (following Salmon); while variation is the most general conceptualization of the patterns of difference-making we can detect in populations where a cause is acting (following Mill). The thought is that we can use these complementary ideas to address the scientific problems.

For example, we can think about how we use complementary evidence in causal inference, tracking information transmission, and combining that with studies of variation in populations. Alternatively, we can think about how measuring variation may help us formulate policy decisions, as might seeking to block possible avenues of information transmission. Having both concepts available assists in describing this, and reasoning well – and they will also be combined with other concepts that have been made more precise in the philosophical literature, such as capacities and mechanisms.

Ultimately, the hope is that sharpening up the reasoning will assist in the conceptual enterprise that lies at the intersection of philosophy and science. And help decide whether to encourage sport, mobile phones, homeopathy and solar panels aboard the mission to Mars!

Recent Comments

There are currently no comments.

Leave a Comment

Your email address will not be published. Required fields are marked *