Back

#SPP2023 Keynote Abigail Marsh:
- Empathy scales repeatedly failed to detect differences between altruists and controls
- altruists were more behaviorally responsive to facial fear, had more activation in the R amygdala (indicating simulation of others' fear), and calcification of the R amygdala dulled the behave. response
- (lots of other evidence)

Claim: altruists aren't more empathic, but feel more connected to (and care more about?) strangers

More on gScholar: https://scholar.google.com/citations?user=1Dj71eIAAAAJ&hl=en&oi=ao

Ny Basil (with Dr. Kate Ritchie and assistants Joanna Villas and Emma Brink) presented data to #SPP2023 showing that "generic generalizations" (like 'dogs bark', 'cars have wheels') can have different impacts on depending on "generic ratings" depending on their scope (e.g., 'dogs all of the world bark', 'the dogs in Turkey bark').

More about Ny Basil: https://cognition.princeton.edu/publications/contributor/vasil-ny

Kate Ritchie on gScholar: https://scholar.google.com/citations?user=-kUUtFEAAAAJ&hl=en&oi=ao

Rebecca Peretz-Lange (and colleagues') shared results and theory about why #essentialism frames can sometimes promote and sometimes mitigate #prejudice.

Dr. Peretz-Lange appeals to studies that included both kids and adults to suggest that "causal discounting" can make the seemingly contradictory results coherent.

You can find/follow the publications from Rebecca's gScholar page: https://scholar.google.com/citations?view_op=list_works&hl=en&hl=en&user=C14WR0gAAAAJ&sortby=pubdate

In the #ArtificialIntelligence session of #SPP2023, Celine Budding presents "Tacit Knowledge and Top-Down Experiments for Explaining How Large Language Models Work"

Dr. Budding proposes "We can conceptualize what #LLMs learn as ...tacit #knowledge".

Implications for how the systems work, allow for interventions on the systems, and provoke hypotheses and experiments.

Find/follow Celine's publications on gScholar: https://scholar.google.com/citations?user=M_sy5A0AAAAJ&hl=en&oi=ao

The 2nd #AI session talk at #SPP2023 was given by Sam Whitman McGrath and Jacob Russin: "Can Deep Learning Inform Explanation in Cognitive Science?"—the slides also credited Roman Feiman, Ellie Pavlick, and Randy O-Reilly.

They take a middle ground view between two existing views—an underrated move in academia?

You can follow this kind of work via Jacob's ResearchGate page: https://www.researchgate.net/profile/Jacob-Russin/research

I could not find a followable webpage for Sam.

Cameron Buckner closes the #AI session at #SPP2023 with "Interventional Methods for Relating Representations in DNNs to Representations in the Brain"

As usual, it was a delightful path from #AI through #computerScience, #PhilosophyOfScience, and #philosophyOfMind!

One teaser take-away: #AI is not (and was not) a stochastic parrot!

Find/follow Cameron's work (including forthcoming book) on gScholar: https://scholar.google.com/citations?user=OXgCldoAAAAJ&hl=en&oi=ao

Outstanding #cogSci/#xPhi #research about #wisdom! (N = 2707) to kick off the "Thinking about Minds" session of #SPP2023:

In 19 #wise characteristics, two factors emerged across #culture:
1. Socio-emotional awareness
2. Reflective thinking

People also rated how much of the characteristics they had compared to certain people (young, middle-aged, old; doctor, politician; etc.).

A 2-factor plot of those relative ratings was *not* universal across 8 regions.

#preprint: https://doi.org/10.31234/osf.io/p9cv4

David Melnikoff presented "Bayesian and Wishful Thinking are Compatible" (a project with Nina Strohminger):

The finding: people felt better than predicted about the prospect of either prosecuting or defending a defendent, no matter which side they were incentivized to defend.

Preprint (pending revision for Nature Human Behavior): https://doi.org/10.31234/osf.io/yhmvw

In the "Thinking about Social Groups" session of #SPP2023, David Kinney presented "Tell Me Your (Cognitive) Budget, and I’ll Tell You What You Value: ..." (in collaboration with Tania Lombrozo).

They tried to design stimuli that could test the degree to which people incorporate a representation of a problem, relevant data, and values.

I didn't see a preprint of this project online, but Dr. Kinney's on gScholar: https://scholar.google.com/citations?user=9gK2xIoAAAAJ&hl=en&oi=ao

Isaac Davis presented "Inferring the Internal Structure of Social Collectives".

Isaac, Yarrow Dunham, and Julian Jara-Ettinger designed social vignettes to test their computational model of how people infer social structure (e.g., hierarchies) from a domain-general statistical learning mechanism and domain-specific social knowledge.

They think 2 experiments "support our account".

Preprint: https://doi.org/10.31234/osf.io/t5hpb

Paul J. Kelly opened the "Reasoning & Explanation" session at #SPP2023 with "Dynamical Models, Scientific Understanding, and Explanatory Unification".

After some objections to accounts of the explanatoriness of explanatory models, Kelly refers to the notion of "model transfer" in which one model can explain multiple phenomena using the same "state space"—an idea that traces to Kitcher (1989).

Paul can be followed on PhilPeople: https://philpeople.org/profiles/paul-j-kelly

Kevin O’Neill presented great #xPhi/#cogSci studies of how people think of causation in cases of double prevention (i.e., A prevents B from preventing C from doing something).

Kevin, Paul Henne, Tadeg Quillien, Thomas Icard, & Felipe De Brigard devised three kinds of cases to understand people's thinking.

Gorgeous plots! But the team seems to admit they deviate from their model.

You can follow Kevin on PhilPapers: https://philpeople.org/profiles/kevin-o-neill

Caren Walker presented "Children Consider Future Learning Goals During Information Search" with/for Liz Lapidow (who was sick, absent).

Puzzle: evidence is mixed about whether kids are "intuitive scientists" or bad at scientific thinking (controlling for variables).

Modified task suggested kids *can be* intuitive scientists when the task dissociates different goals that kids might have during the task.

Paper on gScholar: https://scholar.google.com/scholar?oi=bibs&hl=en&cluster=18137361289857062215

I found some great posters at #SPP2023. So many, that I had to shorten interesting conversations with interesting people to see them all. I'll start posting them below (before returning to the talks, if I have time).

Devin Curry wins the "most interactive" poster (not a real prize). Follow Devin on gScholar: https://scholar.google.com/citations?user=fnhQAmwAAAAJ&hl=en&oi=ao

Jenna Alton et al. found that kids seem to be sensitive to both gender categories *and* gender appearance/expression. I found myself wondering if visual implicit bias tasks control for *both* to overcome measurement error. You can follow Jenna on gScholar: https://scholar.google.com/citations?user=RBQH4d8AAAAJ&hl=en&oi=ao

Pearl Han Li and colleagues found that kids were somewhat persuaded to accept morally questionable actions by weak egoist ("consequentialist") reasons or appeal to authority, but these effects weakened as kids got older (from 3 to 5 years old). You can follow Pearl on gScholar: https://scholar.google.com/citations?user=VCsQ6F8AAAAJ&hl=en&oi=ao

Isaac Handley-Miner found (with Liane Young) that people's epistemic judgments about whether claims are true, known, or believed were sensitive to the *consequences* of the claims. You can find/follow Isaac on gScholar: https://scholar.google.com/citations?user=kPy78bQAAAAJ&hl=en&oi=ao

Hannah Lunkenheimer and colleagues found that kids' view of whether God intervenes on the world seems to depend on how much control they (as kids) have on their lives. Hannah had plenty of promising follow-up research ideas (and has already submitted and received feedback on funding proposals about this). Find Hannah on gScholar: https://scholar.google.com/citations?user=WGxHkRgAAAAJ&hl=en&oi=ao

Madeline ("Gracie") Reinecke et al. found that kids' views about how to treat people, robots, toys, and rocks were best predicted by kids' views of whether those 4 targets had "experience"—perceived "agency' was less predictive. Gracie also has new data, which includes a convenience sample of adults. Find out when this is published from Gracie's gScholar page: https://scholar.google.com/citations?user=AJ_DNT8AAAAJ&hl=en&oi=ao

Claudia Sehl et al. found that people's preferences for simplicity (e.g. simpler explanations) was detected even when they needed to make a decision about how to cause something. Like prior work, the simplicity preference fades when data indicate complexity—"complexity matching". You can follow Claudia on gScholar: https://scholar.google.com/citations?user=JJrQnoUAAAAJ&hl=en&oi=ao

Another #SPP2023 Keynote Elisabeth Camp suggests “Identity Labels [can function] as Interpretive Frames for Building Agency”

Some pushback and requests for clarification in the Q&A.

Find your own clarification from Dr. Camp's gScholar page: https://scholar.google.com/citations?user=RfuoIFYAAAAJ&hl=en&oi=ao

My #SPP2023 poster
1. reflection tests didn't prime intuitions on 10 thought experiments
2. some reflection-philosophy correlations found
3. thought experiments primed reflection?
4. Up to 18x more junk from mTurk than Prolific, CloudResearch, or uGrads

Preprint: https://doi.org/10.31234/osf.io/y8sdm

Melissa Kibbe shared “Function Arithmetic Computations Over Pre-Symbolic Representations of Quantity in Infants and Children” at #SPP2023:

Research with Cheng empirically distinguished nonsymbolic arithmetic (noticing one physical object is placed next to another) from symbolic arithmetic (1+1=2) are algorithmically distinct, which may limit transfer from nonsymbolic arithmetic formal math.

Find/follow Dr. Kibbe on gScholar: https://scholar.google.com/citations?user=NN4GKo8AAAAJ&hl=en

Kevin Dorst then presented “#Polarization is Not (Standard) #Bayesian”

Why not? Because most people violated Bayesian “Martingale updating”.

Why? “Intuitively, we often are (instrospectively) unsure what our probabilities/credences are."

Kevin proposes, “If we want adequate Bayesian models of polarization, they must incorporate introspection failures.”

See photo for the handout.

Ben Rottman: “The Accuracy of Causal Learning and Judgment from Experiences Gathered over Weeks”

Six decently-powered studies found some conditions under which shorter- and longer-term causal learning occurs.

Find/follow Ben on gScholar: https://scholar.google.com/citations?hl=en&user=BMIUsOgAAAAJ&view_op=list_works&sortby=pubdate



Jim Woodward literally talked though “Some Reflections on Actual Causation and Causal Selection”

The slides were walls of text and the talk was Jim’s corresponding and unscripted thoughts.

You can find Jim’s lifetime of work topics like this on gScholar: https://scholar.google.com/citations?user=x0v5vhQAAAAJ&hl=en&oi=ao

In "Rethinking (In)Voluntarism", Dr. Laura Soter points out that voluntarism has focused on the process of appraisal and then turns our attention to a “backend” or “indirect” control (evidence gathering, inquiry, etc.) of belief formation.

Soter argues we have this backend/indirect control and then responds to some objections, thereby reinvigorating doxastic voluntarism.

You can find/follow Laura research on gScholar: https://scholar.google.com/citations?user=LBIbs2YAAAAJ&hl=en&oi=ao

Drew Johnson presents a “Needs-Based Account” (NBA) of attention, develops some desiderata for “need”, considers two objections, and considers future directions in evolutionary psychology, clinical psychology, and epistemology.

You can find/follow Drew on PhilPeople: https://philpeople.org/profiles/drew-johnon



Kristina Krasich & Anna Ghelfi present “Mental Control and Effort Differ Across Different Kinds of Mental Action”

A visual task tested
- relationships between perceived control & effort (positive correlation)
- whether they differ for seemingly passive or active activities(no)
- more

Collaborators: Samuel Murray, Felipe De Brigard, & Joshua Shepard


Dr. Krasich on gScholar: https://scholar.google.com/citations?user=T-vc59sAAAAJ&hl=en&oi=ao

Anna Gelfi on LinkedIn: https://www.linkedin.com/in/annaghelfi26/



Andrew Shtulman's “Reflecting on Possibility: Cognitive Reflection Facilitates the Development of Modal Cognition" replicated and extended finding that kinds think wrong actions are improbable, are improbable events are wrong!

Kids’ reflection test performance predicted kids possibility and permissibility judgments (above and beyond age and executive function).

Follow on gScholar to learn when article is up: https://scholar.google.com/citations?user=Cz8bTrkAAAAJ&hl=en&oi=ao



This paper's in press at the Journal of Experimental Child Psychology: http://sites.oxy.edu/shtulman/documents/2023e.pdf

Dr. Shtulman's research assistants (and co-authors):
Charlotte Harrington
Chloe Hetzel
Josephine Kim
Caroline Palumbo
Teddy Rountree-Shtulman

Emily Liquin presented “The Origins of Information-Seeking Questions”

People
1. re-used previously seen “question templates”
2. even “novel" questions were similar to templates.

This anchoring decreased with age.

Curiously, older kids and adults didn’t ask questions that could reveal more information.

Collaborators: Barron Tsai, Marjorie Rhodes, & Todd Gureckis

Dr. Liquin’s on gScholar: https://scholar.google.com/citations?user=IdymDGQAAAAJ&hl=en&oi=ao

Tamar Kushnir’s #SPP2023 presidential address tried to answer, “When do children become responsible for moral decisions?”

Evidence suggests people’s opinions vary by culture, as do laws, but there’s evidence that kids develop the ability to understand moral aspects of decisions (including that some decisions seem to be moral).

Find/follow Dr. Kushnir on gScholar: https://scholar.google.com/citations?user=TMuSMXoAAAAJ&hl=en

Catching up on the #SPP2023 #preconference on #memory:

Felipe De Brigaard introduced us to the topic and some recent trends before a series of talks ensued.

Find Felipe's work on gScholar: https://scholar.google.com/citations?user=l9gS2joAAAAJ&hl=en&oi=ao

John Anderson presented “The Environmental Basis of Memory”, at #SPP2023’s #preconference on #memory.



Dr. Anderson applied “rational analysis” of human cognition (based on only goals, the environment, and constraints of the cognitive system) to explain memory (with Milson in 1989) and find evidence for this (with Schooler in 1991).

More recent #bigData from #Reddit and #Twitter fit the model.



Dr. Anderson’s career on gScholar: https://scholar.google.com/citations?user=PGcc-RIAAAAJ&hl=en&oi=ao



Katherine Puddifoot presented “Ameliorating Memory”—work with Marina Trakas



The idea was that memory systems may help explain certain biases and prejudices that social scientists observe—a.k.a., #mneumonicInjustice.

This cited philosophers like Miranda Fricker et al.



Katherine Puddifoot on gScholar: https://scholar.google.com/citations?user=UkM8AOMAAAAJ&hl=en&oi=ao

Markus Werning presented “Episodic memory as a predictive process: minimal hippocampal traces as error signals and the role of precision weights” at the #SPP2023 #preconference on #memory

Find Markus’s work on gScholar: https://scholar.google.com/citations?user=V87KRwsAAAAJ&hl=en&oi=ao

Teresa McCormack closed the #SPP2023 #preconference on #memory with “The value of remembering and anticipating experiences: a developmental perspective”

It was—as Teresa put it—dangerously close to an #xPhi talk. It adapted a famous thought experiment (from Derek Parfit?) to test kids’ and adults’ intuitions about how much we care about past, present, or future versions of us.

Follow Dr. McCormack on gScholar: https://scholar.google.com/citations?user=g9T7yn8AAAAJ&hl=en&oi=ao

That wraps up my #SPP2023 experience. Thanks to Chantel and Jennifer for excellent logistics, Edouard Machery for local organizing, and Nadia Chernyak and Sara Aronowitz for co-charing the program!

As usual, I left the SPP thinking, "These are my people". I really hope I can attend again soon!

@ByrdNick thanks for conference reporting - sounds like a great event!