On the Wikipedia page for Mind-body dualism, one of the arguments against dualism is neuroscience.
In some contexts, the decisions that a person makes can be detected up to 10 seconds in advance by means of scanning their brain activity. Furthermore, subjective experiences and covert attitudes can be detected, as can mental imagery.
How can substance dualism survive given these discoveries?
As Peter Reynaert puts it in Reynaert, P. (2015): "Neuroscientific Dystopia: Does Naturalism Commit a Category Mistake?" In: Neuroscience and Critique (pp. 70-86). Routledge (this is a synopsis of his):
My claim that naturalism commits a category-mistake has consequences for assessing neuroscience. Neuroscience is a variant of naturalism which reduces aspects of consciousness and human existence to brain processes. The identification of this mistake has both ontological and methodological consequences. The absurdity of naturalism implies that human consciousness and existence are conceptualized with notions and theories that cannot be applied to them, because they do not belong to the ontological region called nature. Second, and because of this, the scientific methods and theories developed for nature cannot be used to elucidate and understand human existence. Contemporary debates about the (un)reality of the free will for instance, whereby arguments against its existence are allegedly borrowed from neuroscientific research, are simply futile, because they illustrate par excellence the absurdity of the naturalist approach.
This does not necessarily commit him to substance dualism, but this is a basic argument the position often is built upon. I would paraphrase and summarise it as follows: We cannot just discard our phenomenal and normative reality by applying explanations and concepts that are methodically and conceptually tailored to explain physical reality, stating an identity across frameworks.
For example, the quote says "decisions", "subjective experiences" and "covert attitudes" can be detected. Following Reynaert, this is simply not true: neuronal activity can be detected (data) and "decisions", "subjective experiences" and "covert attitudes" are voluntarily and consciously identified (interpreted as being identical) with the physical events detected.
Basically, the neuroscientists that say they can measure mental states have decided that the physical data is the mental state before measuring anything at all since all they are able measure is the occurrence of physical events and the time difference between these occurrences. The identity or non-identity of a physical and a mental event is part of their theory. But the basic hypothesis as such - the identity of neural activity and mental states - is probably not testable at all and cannot become part of the scientific endeavour proper, i.e. it is a contingent premise that determines the interpretational outcomes of the studies.
Thus, we here have a perfect example of a genuinely philosophical question about whether there is an identity or not.
In some sense, dualists do not have to "explain" anything there, they can simply point out that this interpretation is nonsensical (as Reynaert does at length - it is about this being absurd) and that there is no discrepancy to be explained as we are talking about totally different things here.
That being said, substance dualism as an ontological position obviously runs into different problems, i.e. mind-body interaction. It can simply be stated that it works so-and-so, but whatever the explanation ends up to be, it can hardly be proven as the correct interpretation. All there can be is more or less coherent justifications and arguments: It is speculative metaphysics.
Nevertheless, there are many variants of explanations out there that try to explain the temporal discrepancy between physical and mental events that somehow seem to have something to do with each other, though. One possible way to think here is given by Daniel Dennett in Freedom Evolves, p.229:
When you think you're deciding, you're actually just passively watching a sort of delayed internal videotape (the ominous 300-millisecond delay) of the real deciding that happened unconsciously in your brain quite a while before "it occurred to you" to flick.
In other words: You could theorise that there actually is an unconscious decision and that this one's "yours" in a meaningful sense but you perceive it much later on and only attribute/interpret it as your conscious decision happening in that moment.
There are other arguments for mind-body dualism on the "Mind-body dualism" Wikipedia page, but the question is about a specific argument against it coming from neuroscience:
In some contexts, the decisions that a person makes can be detected up to 10 seconds in advance by means of scanning their brain activity. Furthermore, subjective experiences and covert attitudes can be detected, as can mental imagery. This is strong empirical evidence that cognitive processes have a physical basis in the brain.
Alfred Mele addresses the issue from the perspective of free will in Chapter 3 of Free: why science hasn't disproved free will (page 26-39). Mele and the Wikipedia authors seem to be addressing the same research by Soon, et. al.
Mele describes what the experiment asked the participants to do: (page 27)
The study's participants were asked to make many simple decisions while their brain activity was measured using fMRI. Their options were always to press one or the other of two buttons. Nothing hinged on which one they pressed--no reward, no penalty, nothing at all.
Mele notes that the accuracy of the study was 60 percent with 50 percent being pure chance. What might be giving the researchers an edge? Mele suggests: (page 28)
What are the scientists measuring or detecting several seconds before a button press? What is that neural activity associated with? My bet is a slight unconscious bias toward a particular button on the next press.
Mele also objects to generalizing from this experiment picking buttons to all decisions that a person may make: (page 27-8)
My concern now is that this kind of picking may not be very similar to choosing or deciding in situations in which a lot of conscious weighing of reasons--pros and cons--goes into the choice or decision. How similar is the arbitrary picking of a button to a decision to ask one's spouse for a divorce--or to change careers or start a small business--after protracted reflection on reasons for and against that decision. If arbitrary picking is not very similar to these other decisions, claiming that what happens in instances of arbitrary picking also happens in instances of complicated, painstaking decision making is a huge stretch.
Let's consider the OP's question: How can substance dualism survive given these discoveries?
One can use Mele's arguments about possible unconscious bias justifying the 60% success rate over 50% random chance as one way to dismiss the experiment as inconclusive. Because of that explanation, the experiment doesn't offer a serious threat to substance dualism or any other alternate understanding of mind-body interaction.
Also one can reject any generalization of these experimental results to important decision making since the experiment did not cover such important decisions.
Reference
Mele, A. R. (2014). Free: why science hasn't disproved free will. Oxford University Press.
Soon, C. S., Brass, M., Heinze, H. J., & Haynes, J. D. (2008). Unconscious determinants of free decisions in the human brain. Nature neuroscience, 11(5), 543. https://www.nature.com/articles/nn.2112
Wikipedia, "Mind-body dualism" https://en.wikipedia.org/wiki/Mind%E2%80%93body_dualism
The Wikipedia page is wrong in its interpretations and implications of the decision studies.
First, in practical terms, no human decisions could be made 10 seconds ahead. Life moves at a much faster rate than that, and that slow a decision cycle would lead to failure over an over. Try playing Ping-Pong, with a 10 second decision cycle! You will be flailing your paddle before an opponent even serves!
Second -- Libet, the first such experimenter demonstrated that such decisions are not fixed. He showed that we have "free won't" as a minimum. https://www.psychologytoday.com/us/blog/dont-delay/201106/free-wont-it-may-be-all-we-have-or-need
Third -- the 60% success rate of Soon in predicting which button will be pushed, of two, is not the demonstration of determinism, or choice, but of INCLINATION. 100% would show determinism, or choice, 10 seconds ahead. https://www.nature.com/articles/nn.2112
Alfred Mele wrote an excellent book spelling out the problems with over interpreting these experiments: https://global.oup.com/academic/product/free-9780199371624?q=Mele,%20alfred&lang=en&cc=us
But what is actually going on, as SOMETHING is being decided up to 10 seconds ahead? The best answer comes from Dennett's multiple drafts model of consciousness, coupled with Popper's hypothesis forming model. https://wiki.cs.umd.edu/cmsc828p/index.php?title=Multiple_drafts_theory http://www.informationphilosopher.com/solutions/philosophers/popper/natural_selection_and_the_emergence_of_mind.html
Our minds maintain multiple "what ifs", in parallel, and prepare to act on them. Most of this preparation is unconscious. It takes time to organize our neurology to do something properly, so one should EXPECT that any action has some unconscious preparation happening before we consciously choose to implement it.
Dualism is entirely consistent with the Libet and Soon experiments. Dualism assigns the unconscious multiple drafts role to the brain, and the decision role on which draft to implement to the self. The preparations will always preceed the decision -- so the measurement of readiness potential before a decision is made consciously are -- expected -- in a multiple drafts dualist model.
Since there are many answers that seem to be biased towards the assumption, that there is no such thing as the mind which is ontologically different from the brain, I want to add some attitude from my master's thesis (subtitled "The way the informatics enriched philosophy").
To an informatician, "mind-body dualism" sounds like "software-hardware dualism". There are many things in common that may exceed the topic here. I do not think that there is any reasonable scientist out there who says that software does not exist, instead there is only hardware.
From the perspective of informatics, it looks like this: Whenever a process is run, it looks different - depending on the exact hardware, not even on the same hardware (because of the dynamic address assignment). Nevertheless, it is the same software, no matter which hardware. This is at least one indication that software is dependent on, but different from hardware.
In this same sense, mind is dependend on, but different from the brain.
