Object-oriented is a contested term from software language design to ontology. Originally coined to describe the programming language Smalltalk by Alan Kay, one of the key designers of the language, its use has not only migrated to software languages that have very little in common with Smalltalk, but even to philosophical ontology via Graham Harman’s ‘object-oriented ontology’, ostensibly arising from a re-interpretation of Heidegger’s analysis of tool-being in Being and Time, though it requires a simultaneous posit that Heidegger didn’t understand his own analysis. Harman’s re-interpretation, though, is itself subject to seemingly any interpretation that vaguely fits within the scope of ‘Speculative Realism’, while ‘Speculative Realism’ itself appears to be predicated on a near-complete misunderstanding of Hegel, since Hegel’s Speculative Idealism already includes ‘realism’.
My posing the question is not to try to decide for or against these various interpretations, nor to return the term to Kay’s initial intent (something he would probably rather not do himself), but to question whether the term has a determinate meaning or is simply a nebulous phrase to which very different ideas have been attached; and if it does not have a determinate meaning, what the intrinsic implications of the term might be and why it therefore has become an attractive term to hang those ideas on.
I will begin with Alan Kay, though not with his intent in using the phrase, but on the design of the programming language the phrase was intended to describe, insofar as the proper nature of that language itself became clearer after the term was coined:
“Just a gentle reminder that I took some pains at the last OOPSLA to try to remind everyone that [object-orientation] is not only NOT syntax or the class library, it is not even about classes. I’m sorry that I long ago coined the term “objects” for this topic because it gets many people to focus on the lesser idea.
The big idea is “messaging” — that is what the kernel of Smalltalk/Squeak is all about (and it’s something that was never quite completed in our Xerox PARC phase). The Japanese have a small word — ma — for “that which is in between” — perhaps the nearest English equivalent is “interstitial”. The key in making great and growable systems is much more to design how its modules communicate rather than what their internal properties and behaviors should be. Think of the internet — to live, it (a) must allow many kinds of ideas and realizations that are beyond any single standard, and (b) to allow varying degrees of safe interoperability between these ideas. If you focus on just messaging — and realize that a good metasystem can late bind the various 2nd level architectures used in objects — then much of the language-, UI-, and OS based discussions on this thread are quite moot.
This was why I complained at the last OOPSLA that — whereas at PARC we changed Smalltalk constantly, treating it always as a work in progress — when ST hit the larger world, it was pretty much taken as “something just to be learned”, as though it were Pascal or Algol.
I recall also pointing out that it is vitally important not just to have a complete metasystem, but to have fences that help guard the crossing of meta-boundaries. One of the simplest of these was one of the motivations for my original excursions in the late sixties: the realization that assignments are a metalevel change from functions, and therefore should not be dealt with at the same level — this was one of the motivations to encapsulate these kinds of state changes, and not let them be done willy-nilly. I would say that a system that allowed other meta-things to be done in the ordinary course of programming (like changing what inheritance means, or what is an instance) is a bad design. (I believe that systems should allow these things, but the design should be such that there are clear fences that have to be crossed when serious extensions are made.)”
Given that messaging (and related things such as encapsulation of internal properties and behaviours are in fact more crucial to a description of Smalltalk than objects or object-orientation, it doesn’t appear to have an especially determinate meaning for its originator. As a developer familiar with Smalltalk I would myself describe it, in terms of its basic structure of entelechies and messages, as metabolically oriented, if I was forced to include orientation in the description.
To turn on the other hand, to ‘object-oriented ontology’:
“If the human perception of a house or a tree is forever haunted by some hidden surplus in the things that never becomes present, the same is true of the sheer causal interaction between rocks or raindrops. Even inanimate things only unlock each other’s realities to a minimal extent, reducing one another to caricatures…even if rocks are not sentient creatures, they never encounter one another in their deepest being, but only as present-at-hand; it is only Heidegger’s confusion of two distinct senses of the as-structure that prevents this strange result from being accepted.
From this, Harman concludes that the primary site of ontological investigation is objects and relations, instead of the post-Kantian emphasis on the human-world correlate. Moreover, this holds true for all entities, be they human, nonhuman, natural, or artificial.”
Oddly, it appears to be Harman (or at least his interpreter) who has a confusion not only about ‘distinct senses of the as-structure’, since there appears to not be two distinct senses in Harman’s account – one account ‘holds true for all entities’, but also regarding the ‘as structure’ of ‘readiness-to-hand’, where in the tool-analytic the withdrawal of the tool as explicitly present-at-hand in no way implies that it factually disappears and is no longer a relevant problematic.
To explain how withdrawn objects make contact with and relate to one another, Harman submits the theory of vicarious causation, whereby two hypothetical entities meet in the interior of a third entity, existing side-by-side until something occurs to prompt interaction. Harman compares this idea to the classical notion of formal causation, in which forms do not directly touch, but influence one another in a common space “from which all are partly absent.” Causation, says Harman, is always vicarious, asymmetrical, and buffered:
‘Vicarious’ means that objects confront one another only by proxy, through sensual profiles found only on the interior of some other entity. ‘Asymmetrical’ means that the initial confrontation always unfolds between a real object and a sensual one. And ‘buffered’ means that [real objects] do not fuse into [sensual objects], nor [sensual objects] into their sensual neighbors, since all are held at bay through unknown firewalls sustaining the privacy of each. from the asymmetrical and buffered inner life of an object, vicarious connections arise occasionally…giving birth to new objects with their own interior spaces.
Thus, causation entails the connection between a real object residing within the directionality of consciousness, or a unified “intention,” with another real object residing outside of the intention, where the intention itself is also classified as a real object. From here, Harman extrapolates five types of relations between objects:
Containment describes a relation in which the intention “contains” both the real object and sensual object.
Contiguity connotes relations between sensual objects lying side-by-side within an intention, not affecting one another, such that a sensual object’s bystanders can be rearranged without disrupting the object’s identity.
Sincerity characterizes the absorption of a real object by a sensual object, in a manner that “takes seriously” the sensual object without containing or being contiguous to it.
Connection conveys the vicarious generation of intention by real objects indirectly encountering one another.
No relation represents the typical condition of reality, since real objects are incapable of direct interaction and are limited in their causal influence upon and relation to other objects.
This is far from a re-interpretation of Heidegger’s analysis of tool-being, since Heidegger’s notion of a ‘material domain’ or ‘material context’ already implies a verbal notion of material, i.e. as that which signifies, which has significance, and thus always significance-for. However, the more interesting aspect for me is that given Harman has any acquaintance with Heidegger’s work at all (and in fact he does), he would use the term ‘object’ at all, given that it entails a complete rejection of the entire standpoint of Being and Time (not to mention the rest of his output as a philosopher), and a rejection of what specifically grounds the entirety of the analysis of tool-being, not simply one specific interpretation of it.
“Object [Objekt] and thing [Gegenstand] are not the same. All objects are things, but not the other way around; all things are not objects. A danger ensues of holding determinates of objects for determinates of things. Conversely, one is seduced into holding some thingly determinates immediately for determinates of objects and into applying formal points of view to specific observations of things. Since Plato, the blurring of these differences has been disastrous. Now, a phenomenon is neither object nor thing.
Objects, things, and phenomena cannot be placed alongside each other as on a chessboard; rather, this systematization of things also is inappropriate for phenomena.”
“In taking-cognizance-of, what is cognized therein does not have the character of an object but is experienced as significance. A relating, a grouping-together, manifests itself now; therein, a connectedness of significant things that bears a specific logic, a material logic, a structure peculiar to the specific material states of affairs, is formed.”
“The truth of being remains denied as world. In the denial of world there is the injurious neglect of the thing, where the thing is taken as mere object.”
– Heidegger, Martin.
While Harman is welcome to be dismissive of Heidegger’s philosophical stance, why he felt the need to predicate ‘object-oriented ontology’ on a rather thoughtless interpretation of a single passage in Being and Time to begin with is a bit of a mystery. It would have been far easier and have introduced fewer problems into Harman’s work if he simply had begun from the rational-scientific notions of ‘object’ and ‘objectivity’, which are after all far more widely accepted than Heidegger’s complex distinctions between the ready-to-hand, the ‘present-at-hand’, things, works, Dasein, etc. in terms of a fundamental ontology. Since nothing in his work has any real dependency on the tool-analytic, why invoke a thinker whose stance is opposed to those notions from beginning to end? Introducing a thinker who is difficult to deal with (aside from simple dismissal) from the rather generic rational-scientific metaphysics that OOO turns out to be, who would likely never be read Harman’s audience, and whose work is in no way requisite to understanding OOO seems self-defeating at best. Object-oriented ontology appears to be merely a Leibnizian monadology with a lack of appetite. That interaction takes place in ‘some other being’ echoes Leibniz contention that all interaction between monads takes place within God, without the embarrassment directly ascribing interaction to God might cause Harman.
That Harman’s readers miss the fact that Harman’s use of ‘object’ rather than ‘thing’ is intentionally anti-Heideggerian, and thus a sort of ‘argument’ with Heidegger is in some sense important, at least to Harman’s own understanding of his project, can be seen in the following, where an admirer of Harman’s attempts to describe OOO (this is not the first attempt by Mr. Bogost, in fact he incorporated critiques of his initial attempt, so it’s not simply a mistake due to time constraints or some other extrinsic factor):
“Ontology is the philosophical study of existence. Object-oriented ontology (“OOO” for short) puts things at the center of this study. Its proponents contend that nothing has special status, but that everything exists equally–plumbers, cotton, bonobos, DVD players, and sandstone, for example. In contemporary thought, things are usually taken either as the aggregation of ever smaller bits (scientific naturalism) or as constructions of human behavior and society (social relativism). OOO steers a path between the two, drawing attention to things at all scales (from atoms to alpacas, bits to blinis), and pondering their nature and relations with one another as much with ourselves.” – Ian Bogost
‘Contemporary’ for Mr. Bogost appears to be the 18th century, and only the early 18th century at that. However more relevant in this context is the lack of any discrimination between object and thing, the terms are completely interchangeable.
To assist Mr. Bogost should he wish to think about the things he writes, I’ll give a brief correction to the philosophical background of his own description:
- Ontology is the study of the being of beings, not existence. In metaphysical terms, the being of a being refers to a being’s “whatness”, while the existence of a being refers to its bare “thatness”. ‘Whatness” is also seen metaphysically as essence, whether essence is considered to be idea, substance, mathematical description, the union of appetite and perception, noumena, spirit, will, will to power, or (ex)change. The displacement of being by essence was necessary because the ontologies of Plato and Aristotle had already been interpreted ‘metaphysically’, being was already seen as the ‘highest being’, or God, and thus ontology within metaphysics is always an onto-theology. Sartre’s statement that existence precedes essence remains a metaphysical statement, since it reverses the ordering of a metaphysical statement without altering the notions in the statement, while Heidegger’s seemingly related (or at least similar) statement that the essence of human being is existence is no longer a metaphysical statement, since the fundamental meaning of both essence and existence are radically transformed by the statement itself.
- “the aggregation of ever smaller bits” is not ‘scientific naturalism’ but mereological nihilism, as espoused most clearly by Cian Dorr, (mereology is the study of parts and wholes, which Dorr obviates by viewing everything as a simple aggregation of the smallest thing imaginable). In fact Dorr’s ‘mereology’, were it accepted, would make science impossible, since viewing things as a unity of data rather than a simple aggregation is necessary for positing movement, change and every other possible experiential or explanatory conjugate. With a simple aggregation, nothing could be posited of it at all, since in the next moment it would no longer be precisely the same aggregation, and the unity that remains the same while specific conjugates change, of which regularities, probabilities and exceptions can be attributed, would not exist.
- ‘Scientific naturalism’ is nonsensical since philosophical naturalism is inherently extra-scientific (there’s no way that it can be tested or demonstrated either positively or negatively).
- Methodological naturalism, in contrast, is an essential part of every science because science was and remains based in onto-theology (in fact it began as ‘theology of nature’ and only later had the name changed to ‘natural science’). As onto-theological, anything and everything could be explained as simply “because God”. That science can posit reality as mathematical or indeed rational at all depends on the theological notion of Imago Dei, that man is ‘in the likeness of’ God. Finite rationality was intended precisely to allow man to understand reality as the creation of a rational being. ‘Reality as the creation of a rational being’, being a bit unwieldy to use constantly, was given a specific name: nature.
- Seeing things as “constructions of human behaviour and society” is known as social constructivism, not social relativism. Social relativism involves morality as normative, conceding that it’s only normative for a specific society at a specific time.
The use of ‘object-oriented’ for Harman’s ontology appears even more problematic when the second term is considered, given that his classes of possible object relations are precisely not based on being-oriented in any way, they seem to be predicated on the merest form of being-present possible.
I’m tempted to suspect that Harman’s use of the term, along with his introduction of Heidegger’s tool-analytic, is no more than an attempt to make a weak form of Leibniz appear shiny and new. However, to give him the benefit of the doubt, while also not simply accusing Alan Kay of thoughtlessness when introducing the term, I’ll assume that the term itself has some sort of unstated implication for both Harman and Kay, and that their use of it is neither merely accidental nor merely a means of attracting readers.
The question then becomes, what is implied by the term? Since it is not particularly appropriate in either of the main ways it’s factually used, to what might it be appropriate?
On the premise that the meanings words have (and have had) leave a trace even where that meaning is not explicitly thought or intended, we can note a few things about the terms themselves and therefore about their conjoining.
Object doesn’t mean the same as thing, whether one has read Heidegger or not. In common sense terms, object more closely corresponds to a ‘body’, albeit one with its significance removed, or at least severely restricted. Within the rational-scientific context that Harman’s thought gains its sense from, though, the distinction made clear by Bernard Lonergan in the work Insight: A Study in Human Understanding, between ‘things’ and ‘bodies’ will be helpful, as will the meaning of ‘object’ insofar as it is used to describe how both are determined as such. Lonergan begins by differentiating experiential and explanatory constructs, a distinction mirrored by Harman’s distinction between ‘sensuous’ and ‘real’ objects.
To employ an experiential conjugate is to prescind from all aspects of data except some single quality such as ‘red’ or ‘hot.’ To employ an explanatory conjugate is to turn attention away from all directly perceptible aspects and direct it to a non-imaginable term that can be reached only through a series of correlations of correlations of correlations. To speak of a probability is to suppose a process of reasoning that rests, not directly on what is given, nor positively on what can be understood in the given, but indirectly and negatively on what follows from a lack of system in the given. Now the notion of a thing is grounded in an insight that grasps, not relations between data, but a unity, identity, whole in data; and this unity is grasped, not by considering data from any abstractive viewpoint, but by taking them in their concrete individuality and in the totality of their aspects. For if the reader will turn his mind to any object he names a thing, he will find that object to be a unity to which belongs every aspect of every datum within the unity. Thus, the dog Fido is a unity, and to Fido is ascribed a totality of data whether of color or shape, sound or odor, feeling or movement.
Moreover, from this grasp of unity in a concrete totality of data there follow the various characteristics of things. Thus, things are conceived as extended in space, permanent in time, and yet subject to change. They are extended in space; since spatially distinct data pertain to the unity at any given instant. They are permanent in time; since temporally distinct data pertain to the same unity. They are subject to change, since there is some difference between the aggregate of data at one instant and the aggregate of data on the same unity at another instant. Again, things possess properties and are subject to laws and to probabilities. For the very data that, taken concretely, are understood as pertaining to a single thing may also be taken abstractly and so may lead to a grasp of experiential conjugates, explanatory conjugates, and probabilities. Because the data are the same, there results an obvious relation between the insights and between the consequent concepts. This relation is expressed by saying that the conjugates are properties of the thing and that the probabilities regard the occurrence of changes in the thing. Again, the same relation is involved in what is named attribution. The concrete unity embraces a totality of aspects. From various abstractive viewpoints, other notions apart from the notion of the thing are to be reached. But because the same set of aspects yields both the notion of the thing and the other notions, the latter are related to the former, and the relation, considered logically, is named attribution. Thus, to say that Fido is black or that he is a nuisance is to conceive both a unity in a totality of aspects and some aspect out of the totality, and then to attribute the latter to the former.
Lonergan, Bernard. Insight: A Study of Human Understanding
As a unity in data, things are specific to particular areas or patterns of experience, since the experiential data between different areas often have few or no data points in common. Bodies, on the other hand, are specific to one pattern of experience, the biological pattern. By ‘biological pattern of experience’ Lonergan doesn’t intend some sort of experience on the level of a biological system with no higher genera, but the pattern of experience common to conscious animals that do not have socio-political patterns of experience (such as humans and canines do) or other patterns of experience beyond what is necessary for intentional survival.
There may be men that employ the term ‘thing’ to mean what we now name as a ‘body’. Men are not pure intelligences. They are animals; they live largely under the influence of their inter subjectivity; they are guided by a common sense that does not bother to ask nice questions on the meaning of familiar names. Accordingly, it would not be rash to suspect that their usage of the name ‘thing’ does not quite coincide with the account we have given; and it is to follow up this suspicion that in the present section we turn our attention to the notion of a body or, rather, of a ‘body,’ where the quotation marks denote some divergence from the notion to be reached by intelligence and reasonableness.
To begin from a clear-cut instance in which there is no need to suppose either intelligence or reasonableness, let us consider a kitten. It is awake, and its stream of consciousness flows in the biological pattern. Such consciousness is a higher technique for attaining biological ends. It may be described as orientated toward such ends and as anticipating means to the ends. Moreover, the means lie in external situations, and so the anticipation is extroverted. The kitten’s consciousness is directed outwards towards possible opportunities to satisfy appetites. This extroversion is spatial: as it is by the spatial maneuvers of moving its head and limbs that the kitten deals with means to its end, so the means also must be spatial, for otherwise spatial maneuvers would be inept and useless. The extroversion is also temporal: present data are distinct from the memories that enrich them; they are no less distinct from the imagined courses of future action to which they lead. Finally, the extroversion is concerned with the ‘real’: a realistic painting of a saucer of milk might attract a kitten’s attention, make it investigate, sniff, perhaps try to lap; but it could not lead to lapping, and still less to feeling replete; for the kitten, painted milk is not real.
Let us now characterize a ‘body’ as an ‘already out there now real.’ ‘Already’ refers to the orientation and dynamic anticipation of biological consciousness; such consciousness does not create but finds its environment; it finds it as already constituted, already offering opportunities, already issuing challenges. ‘Out’ refers to the extroversion of a consciousness that is aware, not of its own ground, but of objects distinct from itself. ‘There’ and ‘now’ indicate the spatial and temporal determinations of extroverted consciousness. ‘Real,’ finally, is a subdivision within the field of the ‘already out there now’: part of that is mere appearance; but part is real; and its reality consists in its relevance to biological success or failure, pleasure or pain. As the reader will have surmised, the terms ‘body,’ ‘already,’ ‘out,’ ‘there,’ ‘now,’ ‘real’ stand for concepts uttered by an intelligence that is grasping, not intelligent procedure, but a merely biological and non-intelligent response to stimulus. In other words, the point to the preceding paragraphs is not to suggest that a kitten can understand and describe its spontaneity but, on the contrary, to indicate through human concepts the elements in a non-conceptual ‘knowing.’
Lonergan, Bernard. Insight: A Study of Human Understanding
Prior to thinking about either Kay or Harman’s use of the term ‘object’ rather than thing, it may be useful to note that Lonergan uses the term ‘object’ precisely in his descriptions of both things and bodies, and what its meaning is. ‘Object’ means a target of intentionality, whether that of the kitten or the scientists, the man of sound common sense or the socio-political man of action.
Lonergan uses the notion of ‘orientation’ to describe the pattern of experience within which intentionality can operate. The ‘orientation of biological consciousness’ is extroverted, it finds its environment already constituted, and most importantly, it is concerned with the ‘real’.
From other patterns of experience, the notion of ‘real’ might include exactly what biological consciousness discounts as not significant, not relevant, and hence not, for it, real, except perhaps as an obstacle that must be gotten around. However, the implication is that what is thought as ‘real’ is whatever is of significance within, or what is material to, a given pattern of experience.
Lonergan points out, as well, that whatever our specialties, and despite being intelligent and reasonable, the conflation of things and bodies is not purely accidental, a specific intentionality underlies it.
Animals have no epistemological problems. Neither do scientists, as long as they stick to their task of observing, forming hypotheses, and verifying. The perennial source of nonsense is that, after the scientist has verified his hypothesis, he is likely to go a little further and tell the layman what, approximately, scientific reality looks like! Already we have attacked the unverifiable image; but now we can see the origin of the strange urge to foist upon mankind unverifiable images. For both the scientist and the layman, besides being intelligent and reasonable, also are animals. To them as animals, a verified hypothesis is just a jumble of words or symbols. What they want is an elementary knowing of the ‘really real,’ if not through sense, at least by imagination.
Lonergan, Bernard. Insight
In terms of Alan Kay’s original characterization of Smalltalk as ‘object-oriented’, then, we can look at it in terms of how the Smalltalk language has aspects that relate to the implications of the terms ‘object’ and ‘orientation’ in the conjoining. As oriented, what matters to a Smalltalk message sender is that the message is understood and acted on appropriately by its target, the ‘object’ to which the message is sent. Since most messages have a second term, that second term is seen as the ‘object’ with which a Smalltalk message receiver concerns itself with and must understand the message in terms of, and in this second sense it also corresponds naturally with the use of the term ‘object’ in English grammar. Thus, a set of metaphorical relations between the core idea, which Kay had in all likelihood not explicitly recognized yet, and the term that came to mind, can be seen as implicit in the term itself.
In the case of Harman’s object-oriented ontology, things are somewhat more complex. Harman is a more careful and consistent writer than his interpreters, and does have more than a passing familiarity with Heidegger’s work, far beyond the cursory re-interpretation of the tool-analytic that supposedly results in some sort of justification for Harman’s position would indicate. The basic meanings of both terms fit fine with Heidegger’s thinking, but cause Harman’s serious problems. If ‘real’ objects are of no significance to other ‘real’ objects, then the notion of ‘real’ is inverted from the natural notion. This inversion repeats the inversion of Platonism itself, where what appears is not considered real but ‘mere’ appearance, and the ‘real’ lies somehow above or beyond ‘mere’ appearance. Not only is OOO ‘a’ metaphysics, it is the same metaphysics as that of Plato, the thinker from whom idealism got its name. The few modifications made are isomorphic to some of those made to Plato’s metaphysics by Aristotle, although there is far more subtlety and complexity in Aristotle’s understanding than in Harman’s.
Thus, rather than forming an implicit set of metaphorical relations with a core idea that hadn’t yet become clear, the conjunction of the terms object and orientation serves instead as a partial barrier to understanding Harman’s actual position. In effect, the term is used as a barrier to deflect attention from inconvenient particulars of Harman’s work, such as that ‘real objects’ correspond precisely to ‘things’ insofar as they are attributed only explanatory conjugates, which as generalized are purely ideal, and cannot be properly said to exist since only particulars exist; that ‘sensuous objects’ are simply experiential conjugates of particular things, and thus imply an orientation that makes possible the specific intentionality required to grasp experiential conjugates in a particular pattern of experience; finally, and most crucially, that it returns precisely to intentionality and significance within different patterns of experience, which implies the ontological priority of human existence that is key to understanding the tool-analytic, the rest of Being and Time, and Heidegger’s basic philosophical stance throughout his work.
While it remains true that rational-scientific onto-theology is not in itself a religion, it may also be that it is no more than a peculiar offshoot of a particular religion.