07/10/2026
🐕 Roaming with Hounds Diary:
Threat & Fear Response is not the same.
We often say:
“The dog is scared, so it is showing a fear response or fear agression.”
Expressing ourselfs like that is however not evidence based, as such an explanation may be thought of as "common sense" while putting together several different levels of explanation of context within an instant we observe.
The work of Dr Joseph LeDoux and Dr Lisa Feldman Barrett gives us a fascinating way to separate them.
🧠 Threat response is not necessarily fear;
LeDoux has made an important distinction between defensive responses to threat and the conscious experience we call fear.
Mammalian brains have neural systems capable of detecting danger and organising defensive behaviour—freezing, fleeing, defensive & offensive attack which start with visceral actions, so changes in heart rate, respiration, endocrine expression that lead to changes in arousal, valence, orienting & attention, and other protective actions / behaviour outputs for the purpose of survival.
These defensive action patterns can operate without requiring the conscious experience of fear.
LeDoux therefore argues that we shouldn't simply equate this threat circuitry as defensive behaviour that is activated from "fear", as though they are all the same thing. (LeDoux, 2012 ¹, LeDoux 2014 ², LeDoux et al, 2016 ³, ⁴ LeDoux, 2018, ⁵ LeDoux, 2022)
The mechanisms for these concepts are not the same anatomically / physically but very intertwined.
This is where Feldman Barrett's work on the theory of constructed emotion (Feldman Barrett, 2017 ⁶) becomes particularly interesting.
🧩 So what does “constructed” actually mean?
Construction does not mean imaginary. Construction relates to the physical activity that builds everything from unconcious electro-chemical action patterns to concious feelings (affect), emotions, memory, thought, meaning making / categorisation, sensory experiences like experiencing colour, the breeze tingling over your exposed skin, the sound of a song that may become an "earworm" for some time during the day, etc.
A brain does not simply receive an objective sensory signal and then attach a pre-existing label to it.
In Feldman Barrett and Earl Miller's 2026 paper, “Categorization is ‘baked’ into the brain,”⁷ they argue that categorisation occurs throughout neural processing, from very early in the bidirectional output & input processing streams.
These compression and decompression gradients are anatomically found on every level of neuronal structure. From many small, not well connected neurons to larger and larger heavily connected neurons.
The brain is continuously using predicted action patterns / previous experience to organise incoming signals into meaningful categories.
In other words:
The brain doesn't first perceive the world and then decide what it means.
Meaning making is the active process of electro-chemical compression and decompression gradients interacting and constructing. These active processes lead to perception itself.
Barrett's theory of constructed emotion applies this active construction principle to emotion.
The brain's action patterns are constructions, what we conciously categorize as past experience, concepts like emotions that incompass the body's changing internal state and the current context of ones own body moving in relation to the environment it was prepared for.
When an emotion concept is constructed, the result is an instance of emotion.
🐕 And this creates a really interesting problem when we watch a dog.
Imagine a greyhound suddenly freezes when another dog approaches.
The dog is processing sensory information alongside the predicted action patterns, learned categories, so its own historical past actions and the changing bodily states within the environmental changes.
It construct efficient actions and motor behaviours needed relating to the situation in a way that makes these particular action also conciously meaningful for example:
freeze → monitor → withdraw→meaning
or
freeze → assess → approach→meaning
▶️or something entirely different.
But we, as the observer, receive another stream of sensory information.
We see the dog's posture, movement, facial expression, ears, tail, distance, context and previous behaviour.
And our brain also categorises using our past history, while moving in the niche we are prepared for.
We might construct:
“fear.”
Someone else might construct:
“uncertainty.”
Another person:
“frustration.”
Another:
“defensive behaviour.”
And another:
“the dog wants to escape.”.....
The behaviour hasn't necessarily changed.
The concious meaning we construct from it has.
Feldman Barrett's earlier work makes precisely this broader point: emotion categories are not simply natural kinds sitting out there in the world waiting to be detected. Our brains use concepts and past experience to categorise sensory information and give it meaning.
🔄 So there are TWO constructions occurring in one interaction, one instance.
The dog's construction
Sensory information + bodily state + previous experience / predictions / actions → meaning
The observer's construction
Dog's behaviour + context + observer's previous experience / predictions / actions + concepts → interpretation / meaning.
And those two constructions do not have to match.
They may vary enormously.
The same instant of interaction can therefore contain:
one behaving animal + one observer + two very different constructions of what is happening.
That is incredibly important when we use words such as fear.
Because saying:
“The dog is afraid”
may actually be telling us more about our categorisation of the dog's behaviour than about the dog's subjective experience we have NO direct access to. Behaviour output is not a direct access to feeling, its one clue.
🐾 So perhaps we should begin somewhere else.
Instead of immediately asking:
“Is the dog fearful?”
we can first ask:
What is the dog doing?
Is this defensive or offensive?
Is there a threat response?
Is a goal being restricted?
What action appears to be being prepared?
What predictions might make that action useful?
What experience could change the dog's future categorisation of the situation?
This brings together circuitry & their mechanics, prediction, categorisation and behaviour without pretending that one observable behaviour gives us direct access to an animal's subjective emotional experience.
And perhaps that is the real lesson:
We don't just observe behaviour.
We categorise it.
And the dog doesn't simply receive the world.
It is continually constructing a meaningful world from its own history / predictive actions, bodily state in relation to the world it moves in.
That means when a dog and a human meet, there are two predictive brains constructing meaning from the same moment—each from its own history, each constructing its own concepts, categories, the meaning of that shared instance.
🐾
¹ LeDoux, J. E. (2012). Rethinking the emotional brain. Neuron, 73(4), 653–676.
² "Coming to terms with fear"
Joseph E LeDoux. Proc Natl Acad Sci U S A. 2014.
³ LeDoux, J. E., & Pine, D. S. (2016). Using Neuroscience to Help Understand Fear and Anxiety: A Two-System Framework. American Journal of Psychiatry, 173(11), 1083–1093.
⁴ Surviving threats: neural circuit and computational implications of a new taxonomy of defensive behaviour
Joseph LeDoux et al. Nat Rev Neurosci. 2018 May.
⁵ LeDoux, J. E. (2022). As soon as there was life, there was danger: the deep history of survival behaviours and the shallower history of consciousness. Philosophical Transactions of the Royal Society B.
⁶ Barrett, L. F. (2017). The theory of constructed emotion: An active inference account of interoception and categorization. Social Cognitive and Affective Neuroscience, 12(1), 1–23. doi:10.1093/scan/nsw154.
⁷ Barrett, L. F., & Miller, E. K. (2026). Categorization is ‘baked’ into the brain. Nature Reviews Neuroscience, 27, 435–456. doi:10.1038/s41583-026-01036-2
