Discussion

If you want your statement about Retrodictive Science posted here then please email  philandrews@pre-historic.com and it will be posted here. 

There are distinctions between predictive experimental science and the historical (retrodictive) sciences. 

Historical science narratives use present data to infer past events.  The Oxford Dictionary of Philosophy (2024) defines retrodiction as: “The hypothesis that some event happened in the past, as opposed to the prediction that an event will happen in the future.” 

Jeffares (2008, 470) says: “The historical sciences seem to make retro-dictions— claims about the past... The historical sciences cannot directly confirm their hypotheses about the past with observations due to the lack of access to the past. They cannot confirm their hypotheses with contemporary observations because they are unique hypotheses about particular times or places. The result is a problem of confirmation. With no ability to observe their objects of enquiry, to repeat observations, or to intervene on processes, there is seemingly no way directly to confirm hypotheses. What I want to show in this paper is that this view of the historical sciences I have presented overlooks their reliance on regularities. Because the historical sciences do utilise regularities, even when investigating one off, singular events, they gain access to the confirmatory apparatus of the experimental sciences. They have the means to test hypotheses about the past... The historical sciences also seek regularities in the world, and have to do so in order to secure their claims about the past... The best way to understand the historical sciences is to see them deploying well understood regularities." 

Also, Cleland (2011, p 552) states: “Practices of the historical natural sciences do not seem to closely resemble those of stereotypical experimental science because historical natural science targets long-past, token events, upon which controlled experiments cannot be conducted." 

Interestingly, narratives about the pre-historic era cannot include experiments because experiments are evaluations of predictions about the future, therefore narratives about the pre-historic era can only use the established regularities of nature (present data) to infer past events. 

Jeffares (2008, 472) says: “[retrodictive] theories make quite explicit claims; they effectively make predictions about what we should see in the record of the past.”

Cleland (2011, p 567) also comments that: “The basic idea behind narrative explanation is to construct a story—a coherent, intuitively continuous, causal sequence of events centering on a precipitating event and culminating in the phenomena (traces) in need of explanation” that we see in the world around us now.

Grim et al (2013, p 2369) say: “In cases in which simulation is used for retrodiction, it is the input conditions that are read for new information. If the output conditions correspond to the current state of the world, and if the simulation’s mechanism plausibly corresponds to ways in which we know the world to work, the input conditions indicate a possible previous state of the world.” 

Thus researchers should do 2 things:

1. Researchers should evaluate the internal consistency of retrodictive narratives about the pre-historic era by theoretically demonstrating how the chain of events conform to the established regularities of nature;  and 

2. Researchers should show that their narratives’ predictions concerning the present align with the world at present and the artifacts we find in it.

Employment of this feedback loop inductive triangulation  (Evans and Thébault (2020), Wylie (2020), Chapman and Wylie (2016), Grim et al (2013) method will fine-tune narratives and increase reliability. 

Generally, narratives have not been required to demonstrate these strategies which means they could be unreliable.  As Odenwald (2022, p 23) says about his narrative of the origin of the universe: “Of course, this entire story is highly speculative, even fanciful.” 

This unreliability is only possible because retrodictive narratives about the pre-historic era have little bearing on the safety or quality of human life, since such narratives are not used in Engineering, Medicine etc. Interestingly, this means that narratives  proposing pre-historic events can be completely wrong and will not cause safety harm to people. (Note that this is different to human history accounts and scientific test result reports, which must be correct because they are applied in Engineering, Medicine etc.) 

Use of the 2 steps noted above in a feedback loop will help increase the reliability of the historical sciences (like in predictive science) because the method requires researchers to theoretically show how there narratives align with the established regularities of nature.

Thankfully Johansson (2016) does address the topic of the historical sciences in his textbook Philosophy of Science for Scientists.

Sadly the other popular university textbooks on Philosophy of Science do not address or acknowledge the historical sciences at all. The textbooks reviewed were by Godfrey-Smith (2021), Griesmaier and Lockwood (2022), Staley (2014), DeWitt (2018), Chalmers (2013), Curd, Cover and Pincock (2012), Ladyman (2012), and Massimi (2014).

Do you think Philosophy of Science textbooks should include at least a reference to the historical sciences, send your reasoning to philandrews@pre-historic.com