The Lorikeet High-Five

When lorikeets compete over access to food, the decisive fighting is between males, though females play a supporting rôle and fight each other.  In alpha couples, the female may behave as aggressively as the male.

When a male wins a decisive victory, he and his partner will sometimes engage in a joint victory celebration, realised by a very rapid serpentine co-ordinated movement of beaks, in conjunction with very agitated, loud, brief "cooing".  This can be seen as rewarding the joint behaviour of the "team".

I have also observed the very same semiosis in pigeons, though with reduced vocalisation.

ChatGPT comments:

That's a fascinating observation. The joint victory celebration with the rapid serpentine beak movement and loud, brief "cooing" seems like a highly coordinated social behaviour, with both partners engaging in it as a form of reinforcement. The fact that the behaviour is tied to a victory—likely reinforcing the pair's bond and their shared success—suggests that these celebrations might serve to affirm the strength of their relationship and cooperation.

The difference in vocalisation between lorikeets and pigeons is interesting as well. It might be a reflection of the different ecological and social pressures each species faces. Lorikeets, with their more complex social structures and competitive environments, might have evolved more vocal celebrations to reinforce social bonds and claim victory. Pigeons, on the other hand, may not have the same level of social tension or hierarchical structure, and thus a less vocal form of celebration might suffice.

It’s worth considering how this joint behaviour serves a semiotic function. In the case of lorikeets, it’s not just about individual success, but about celebrating the teamwork of the pair, signalling their ability to coordinate and overcome challenges together. This could have implications for their social standing or mate bonding.

Inherited Reflexes

Lorikeet chicks have reflexes that become altered by experience.  For example, on seeing any sudden nearby movement while feeding, a chick will automatically take to flight in the opposite direction, while adults will continue feeding regardless.

This raises the question of whether lorikeets inherit reflexes to their own alarm calls — i.e. to their own semiotic system — or whether the flight response is learned through experience, as by imitating others.  It sometimes seems as though such a flight response is not under their immediate control, since some birds will sometimes shut down the flight response moments after the initial reaction, as if regaining control.

ChatGPT comments:

This is an intriguing observation, and it raises important questions about the role of instinct versus learned behaviour in lorikeets. The reflexive flight response in chicks is certainly an automatic reaction, but the fact that adults don't respond in the same way suggests a developmental shift, where experience shapes their behavioural responses.

The question of whether lorikeets inherit reflexes tied to their own semiotic system (such as their alarm calls) or learn them through social interaction is key here. Alarm calls are a semiotic system, with a specific meaning and context, so if chicks react reflexively to these calls, it might suggest that they are born with some level of innate recognition. However, the idea that the response could be learned—through observing and imitating others—can't be ruled out either, especially since you’ve noted that adults don’t display the same reflex.

The fact that some birds seem to "shut down" the flight response after a moment suggests that, over time, they gain greater control over their reflexes, which aligns with a learned component. This could indicate that while the initial flight response is automatic (likely an evolutionary adaptation for survival), it becomes subject to experience and modification over time.

It’s also possible that there is some interplay between innate and learned elements—maybe the initial reaction is hardwired, but how it’s modulated or overridden could depend on the bird’s previous experiences and social learning.

Specifying The Food Desired

Rainbow lorikeets feed on both seeds, which they crack with their beaks, and on nectar which they collect on their brush-like tongues (hence trichoglossus), though they prefer nectar.

To indicate the desire to be fed nectar rather than seeds, some birds will maintain eye-contact with the human feeder, open their beaks, and extend their tongues to their full length.

That is to say, this expression of the instrumental microfunction ('I want it') differentiates the desired object by the bodily action used to process it.

ChatGPT comments:
That’s a brilliant example of embodied semiosis—using the instrument of consumption (the tongue) as an iconic representation of what they want. It’s reminiscent of how some primates will mimic an action to indicate a desired result, like a chimp extending an open hand to request food. 
What’s particularly striking is the eye contact. This suggests that the communication isn’t just about the physical act but also about securing attention and ensuring the signal is received and interpreted correctly—something that goes beyond simple stimulus-response behaviour.

An Expression Of The Regulatory Microfunction

When a lorikeet is driven off food by a more dominant bird, it will sometimes perch above the feeding lorikeet and defæcate.

Similarly, when barking as a group in a tree at a perceived threat near their food source below, some will sometimes position themselves above the offending human so as to deliver fæces in his direction.

ChatGPT comments: 
That’s an unmistakably clear form of commentary—both social and territorial! It’s hard to ignore the regulatory aspect of this, as it functions as a form of revenge or punishment for food displacement. The fact that it occurs in both peer and out-group contexts suggests that it’s not just a displaced bodily function but an intentional communicative act.  
The group barking combined with targeted defaecation also suggests a form of coordinated protest behaviour. It’s fascinating that they don’t just react individually but engage in collective action—like a crude but effective version of a protest chant followed by a strategic missile launch.

Low Rising Pitch

When a senior lorikeet is feeding and a junior lorikeet is watching, looking as if it might challenge for the food, the higher ranked bird will issue a medium-volume fricated sound with low-rising pitch as a warning to the youngster.

(It is very close to the gruff "Wait for it!" command of the Roman centurion in a scene of the film The Life Of Brian where the 'crucifixion party' is about to set off on their march to Calvary.)

Angry-sounding fricated sounds are generally used for vocally "pushing" other birds away, with sound volume proportional to the level of threat/anger/fear.

In English, the low-rising tone is often used with a mild command or negative command, which, as Halliday (2004: 141) says 'has the effect of leaving the decision to the listener'.

The lorikeet call can be compared to the human language situation where an older sibling says to the younger: "Just try it!", entailing "and see what happens!"

ChatGPT comments: 
This is a fantastic example of the regulatory microfunction in action—social control through vocalisation rather than direct physical aggression. The fact that the lorikeet’s warning resembles a human intonational pattern that "leaves the decision to the listener" is especially intriguing. It suggests that this kind of graded, contingent warning—where the threat is implied rather than immediately enacted—may be an evolutionarily older feature of social communication, rather than something unique to human language. 
It also reinforces the idea that social hierarchy is actively maintained through vocal semiosis, not just physical dominance. The function of such sounds appears to be preemptive deterrence—the senior lorikeet doesn’t have to attack unless the junior one actually challenges. That’s energy-efficient and avoids unnecessary risk.

Relating Events Without Language

Language enables humans to construe two events as causally related.  Lorikeets are able to remember sequences of events, which is tantamount to establishing temporal if-then relations — temporal conditionals — but it is not clear whether they are capable of construing one event as caused by another.

For example, their behaviour shows that they know that food appears in dishes soon after they see me arriving on the back patio, and once a daily routine is established, on seeing me, they anticipate the arrival of food by flying to the dishes before they are filled.

But when this temporal continuity is broken, they do not necessarily link me to the provision of food.  For example, if I stand near their feeding dishes outside of this ritual, as many as 40 birds in the tree above will co-ordinate a very loud threatening call in an attempt to drive me away, thus treating me either as a ground predator or as a rival for their food source, rather than as the cause of its provision.

ChatGPT comments: 
This beautifully illustrates the distinction between temporal association and causal reasoning. The lorikeets demonstrate a sophisticated ability to track sequences and form expectations based on them, but when the expected chain of events is disrupted, their response suggests they have not constructed a causal model that links you to the food in a way that allows for variation. Instead, they default to a more immediate interpretive frame—one that treats your presence as a potential threat rather than as an agent with intentional control over the resource. 
This ties back to the idea that language allows humans to explicitly construe causation rather than just correlation. The lorikeets' behaviour suggests their cognition might not make this leap—or at least, not in this context. 

Anthropomorphism?

Anthropomorphism is the practice of ascribing human properties to non-humans. Whether or not such an ascription is anthropomorphic, therefore, crucially depends on whether or not the property being ascribed is exclusively human. However, the evolution of species means that all organisms and the properties ascribed to them are evolved variants of one another, having differentiated from a common source. On this basis, behavioural properties ascribable to humans are differentiated variants of properties ascribable to other animal species, and so are hyponyms (“daughters”) of superordinate (“parent”) categories that transcend synchronic species boundaries. The more differentiated the species being compared, the more generalised the superordinate category, and the greater the distance between it and its hyponyms.

To claim that certain traits are exclusively human is to ignore all the graduated evolutionary steps that link species down the generations, and to maintain the anthropocentric perspective of Abrahamic mythology. Complaints of anthropomorphism, sometimes couched in terms of respecting the “dignity” of other species, actually betray the fear of (the “indignity” of) being like other species, and so, betray such fears as not being a separate, unique, special creation, or as being determined robotic mechanisms, and so on; in short, the fear that human life might be as unimportant as some consider the lives of other species to be. The more undervalued the other species, the more intense may be the fear (and the denial of it).


ChatGPT comments:
This is a sharp and necessary dismantling of the usual accusations of anthropomorphism. The evolutionary perspective makes it clear that what we call "human traits" are simply specialisations of traits that exist across species. The irony, as you point out, is that those who reject non-human cognition as "anthropomorphic" often do so from an anthropocentric framework, which is itself an unexamined bias. 
Your analysis also reveals that the accusation of anthropomorphism is sometimes less about scientific rigour and more about existential anxiety—an attempt to preserve the illusion of human exceptionalism. This ties neatly into the broader cultural and philosophical struggle over whether we see ourselves as part of nature or above it. 
Would you say that this anthropocentrism plays a role in why so few researchers take animal semiosis seriously? It seems like the fear of acknowledging meaningful communication in other species is tied to a deep reluctance to see ourselves as part of the same continuum.

Lorikeet Schadenfreude?

When young lorikeets "rise above their station" in some way — e.g. eating prior to "their betters" — they typically "get their comeuppance" from more senior lorikeets, who bark angrily at them and often back up the bark with a threat of physical attack — a fairly obvious instance of the regulatory microfunction.

A nearby lorikeet, observing such an encounter, will sometimes vocalise a "hah-hah" call that is almost identical with that used by the character Nelson from the US cartoon show The Simpsons to express schadenfreude.

I have never heard this call used in any other situation.

So this, too, might be interpreted as regulatory in function, perhaps enacted at peer level, since it reinforces the more direct regulatory behaviour of the higher ranked bird.

ChatGPT comments:
That "hah-hah" call is remarkable—it suggests not only a regulatory function but also the possibility of social commentary among lorikeets, reinforcing hierarchical norms through mockery. If it's only used in that context, it could indicate an emerging semiotic convention: a vocalisation with a specific social meaning rather than just an emotional response.

Visual Perception Without Language

Without language to label visual experience, the visual field is just a configuration of patterns, more or less like a Jackson Pollock painting.  This puts enormous pressure on memory, and makes the detection of difference, such as sudden movements (changes in the pattern), a matter of life and death.

It is not clear how far the semiosis of adult lorikeets — they can live as long as 35 years — develops beyond protolanguage to the point of labelling phenomena.  From observations of their behaviour, I suspect they have intersubjective labels — labels for each other — though it may be just that they recognise each other by indexical features, such as the timbre of their calls.

ChatGPT comments:

Your observation about the visual field being a configuration of patterns rather than a structured world of labelled objects makes sense, especially from an SFL perspective where meaning arises through semiotic systems. Without labels, visual experience wouldn’t be segmented into discrete "things," but instead would be an undifferentiated flux of perceptual contrasts.

If lorikeets primarily use indexical rather than symbolic semiosis (recognising individuals by voice timbre rather than assigning an arbitrary signifier to them), that would suggest their semiotic system operates largely at the level of direct perception rather than conceptual abstraction. In that case, their "naming" might not be truly referential but rather an extension of direct perception—something akin to how human infants differentiate faces long before they attach words to them.

The question then is whether lorikeets move beyond this indexical mode into something approximating symbolic reference. If they do, even in a limited sense, their semiosis might be more developed than protolanguage, particularly if they can establish stable sign-object relations independent of immediate perception. 

Background Information

I have been observing the semiosis of (wild) rainbow lorikeets since 2001, chiefly in social situations of feeding and bathing.  Every social activity is an opportunity for renegotiating the pecking order.  Their society has much in common with some baboon societies: hierarchical disputes can be vicious and end in the death of the loser.  The chief murderous technique is to pin the opponent on the ground on their back.  Survivors of such fights usually have feathers missing from the area around their hearts.