Varieties of Subjectivity

Recent work in analytic epistemology, philosophy of language and of science has brought back forward the distinction between the objective and the subjective (MacFarlaneforthcoming), (Gallison & Daston 2007), (Wright 2003), (Kölbel 2003, 2000), (Daston2001), (Searle 1995). As Searle (1995, 8) has clearly stated, the distinction works at different levels and as (Daston 2001) has also stressed, these senses have historically evolved. In the following section, I will offer a taxonomy of senses and applications of the distinction, trying to make justice to the contemporary literature on the topic, but also necessarily diverging from all previous accounts of this polysemy. 

My starting point is Daston’s claim that our current understanding of the distinction is based on the identification of particular subjective factors, like perspectives, linguistic conventions, psychological architecture, etc. These facts range from (i) the most personal and temporary, like our preferences, attitudes, feelings and perspectives, to (ii) those we share with other members of identifiable social-groups, like the linguistic conventions of a common language and other historical factors, and even (iii) those we share with others because of some common biological properties, like those we may share with people of our same sex or health conditions (Lloyd 1995). Whatever depends on any of these factors is broadly termed “subjective”, and what is not subjective is called “objective”. In order to differentiate between the aforementioned three different sources of subjectivity, it may be useful to talk about “private”, “social” and “psychological” subjectivity respectively (Swoyer 2008). Once this distinction is in place, it is easy to notice that some philosophers draw broader or narrower limits around the subjective. In a very narrow sense, anything besides private subjectivity is considered objective (this, for example, is Searle’s position regarding what he calls “epistemic subjectivity” in 1995). For others, social factors are clearly as subjective as private ones, but exclude psychological factors, especially those that are species-specific. (Stitch 1990)  For them, for example, (at least some) psychological phenomena may be as objective as material ones, even if they are strongly dependant on our biological makeup. In contrast, others, most notably Frege (1884) and many other early analytic philosophers clearly took a strong view of objectivity, where psychological factors were deemed too subjective. (Jacquette 2003) 

On the other hand, “Objective” and “subjective” are adjectives that distinctly apply to entities of different categories: knowledge, judgments, propositions, objects and concepts. At the ontological level, subjective and objective are different modes of existence (Searle 1995, 8). An entity (object or concept) is subjective if its existence depends on one or another subjective factor. Toothaches, baseball teams and colors are all subjective entities, yet their subjective nature is radically different. Pains are private, baseball teams are social (or “institutional” to use Searle’s term), and colors are psychological. They would not exist, were it not for our psychological makeup, our personal subjective perspective and our sports institutions. 

Subjective Truth 

At the propositional level, a proposition is subjective if whatever makes it true (or whatever determines whether it is true or false) includes or depends on subjective factors. Once again, it may be fruitful to distinguish at this level different kinds of subjective truth: private, social and psychological. Propositions that have determinate truth values independently of any subjective (private, social or psychological) are objective. At this level, propositions like “Wheat Oats are delicious with milk”, “Austin is the capital of Texas” and “The Sky is Blue” are all subjective truths. The facts that make them true are not objective. Some are made true by private facts, others by social facts and finally some may be made true by psychological facts (Nagel 1974). Of course, some times people tend to use the term “fact” to refer only to objective facts (Kripke 1982) and not to any subjective factors that make these other kinds of truths true. So, when people talk aboutfacts, they often mean objective facts, unless otherwise stated. 

Relativism and Context-Sensitivity 

Truths of this kind are also called “relative”, because their truth-value is not absolute, but sensitive to changes in perspective, context, etc. Instead of having a determinate truth-value, their truth-value is relative to a set of parameters that may vary among individuals, moments in time, social groups or even psychological features. Recent philosophy of language has exploited this feature of subjectivity to devise a test for relativity: so-called “context-shifting arguments”. The main idea behind these arguments is that if the truth-value of a sentence shifts in response to certain changes in the context of utterance or evaluation of the sentence, then the proposition may be subjective or relative (Cappelen andLepore 2003). Of course, not any sensitivity to context is enough to talk about subjective or relative truth. Sensitivity to those features of the context associated to traditional indexical expressions like “I”, “here”, etc. are not considered as evidence of subjectivity, because such features themselves are not subjective. Otherwise, even the proposition expressed by “That tree over there is an Oak” would be considered subjective, which is nonsense. So, for a proposition to be subjective, it must be sensitive, not to any change in the context, but only to changes in its subjective features. Thus, only if the truth value of the proposition expressed by a sentence is relative to certain subjective features of the context of utterance, including personal features of the speaker (or hearer), its historical and social context or its biological makeup, then it may change truth values if it is uttered in different contexts. 

As a corollary, just as subjectivity manifests as context-sensitivity, objectivity manifests as invariability. In other words, just as every sentence that expresses a subjective proposition is context-sensitive, every objective proposition is expressible in a context-invariant proposition, i.e. one whose truth value remains stable across contexts (Lycan1996).  But of course, as stated above, not every context-sensitive sentence expresses a subjective proposition and not every context-invariant sentence expresses an objective proposition. “Axel Barceló is in indescribable pain at 2:19 pm on the 13th of May, 2009” is an invariant sentence, yet expresses a subjective proposition. 

Genuine Subjectivity 

Mandik (1998), following a similar point in Searle (1995), has urged us to further refine the above definition of subjective truth (what both Mandik and Searle call “epistemological subjectivity”) to exclude those propositions that are subjective only because of the subjective mode of existence of the entities they are about, and demand that genuine subjective truth stem from the subjective nature of its predicates or the properties andrelations they refer to. After all, any proposition that essentially refers to a subjective entity has truth-value only in those contexts relative to which it exists, and lacks truth-value in the rest of them. This would be enough to make it subjective, if we do not add Mandik’s further condition. In order to understand the motivation behind Mandik’s condition compare, for example, the proposition that New York is to the north from here with the proposition that New York is fun. In both cases, we are talking about the same institutional entity, New York, and therefore, in both cases we have subjective propositions (for they both would have no truth-value had things been such that New York was never founded or it had disappeared by now, for example). However, in the second case, there is a second and more fundamental source of subjectivity: the property of being fun In the first case, we say something objective (the location) of something subjective (New York), while in the second case we say something subjective (that it is fun) about the same subjective thing (New York). For Searle, this means that only the second proposition is genuinely subjective. To acknowledge Searle’s distinction, from now on we may talk of ontologicalsubjectivity to refer to cases of subjective truth that depends solely on the subjective mode of existence of their entities, and genuine or strong subjectivity to talk about those subjective propositions that are not ontological. Thus, a proposition like “Guacamole is tasty” is strongly subjective, while a proposition like “Guacamole has avocado in it” is not, it is only ontologically subjective. Finally also, if T is a term that refers to subjective entities, then the sentence “Ts exist” expresses an ontological subjective proposition: true relative to those contexts where Ts exist and false relative to those where they do not. 

Consensus, Agreement and Intersubjectivity 

An important phenomenon associated to subjective truth is so-called “faultless disagreement” (Kölbel 2003). This sort of subjectivity makes it is possible for two parties to disagree regarding the truth value of a given proposition, not because of any substantial fault on the part of the participants (or, to be more precise, no fault in their inquiry on the truth of such proposition), but because of the matter under disagreement itself. If the parties in disagreement do not share the subjective features that determine the truth-value of the proposition, then each one of them they may faultlessly take it to be true while the other takes it to be false (or lacking truth-value).  

Even though the term comes from the work of Kölbel, this way of cashing out epistemic objectivity originates in the pragmatism of Charles Peirce (1877), and was recently updated by Crispin Wright (1992). Like Peirce before them, Wright and Kölbelconceive of objectivity as the end result of an idealized rational inquiry, i.e. as agreement between ideal rational inquirers. Theories of objectivity of this kind are called consensus,intersubjective or agreement theories, in contrast to so-called mirroring or correspondancetheories of objectivity that hold that the objectivity or subjectivity of propositions depends primarily on the objectivity or subjectivity of what those propositions are about. (Rorty1979, Gauker 1995, ) 

Epistemic Subjectivity 

Besides entities and propositions, there is also meaningful talk of subjective or objectivejudgments or beliefs. For someone’s belief to be subjective, at least one of the grounds upon which the belief is based must subjective, otherwise the belief is objective. If, for example, I base my judgement of the taste of a cigar on subjective aspects of my personal experience smoking it, then my judgement may rightfully called subjective. Notice, however, that I may make a subjective judgement on an objective proposition. I may ground my judgment of, for example, whether my parent’s place is farther from my home than my office at the university (which is clearly an objective matter of fact) on my subjective appreciation of how longer a drive to one is in comparison with a drive to the other. 

Just as we can talk about subjective and objective judgment, we can talk about subjective and objective warrant or justification, if the grounds for belief or judgment are warrant or justification conferring. It is quite controversial, however, whether it makes sense to talk about subjective knowledge or not. For a subject S to subjectively know a proposition p, S would need have subjective grounds for his belief in p that are strong enough to qualify as knowledge. For example, I may know subjectively what it is like to be me or to feel the things I do (Nagel 1974), or I may know subjectively how red things look (Jackson 1982). However, for some philosophers subjective grounds can never be strong enough to qualify as knowledge. From a physicalist perspeitve, for example, if an agent knows a proposition, all his grounds for it must be objective. (Dennett 1991) 

Just as there is a phenomenon of faultless disagreement at the level of truth, we may define an analogue faultless debate, when two parties may not be able to dissolve a dispute or disagreement, not because of any substantial fault on the part of the participants or of the matter under discussion, but because the parties may not share the subjective features that ground their judgments. Unlike faultless disagreement, in cases of faultless debate, one may be warranted, justified or even know that a proposition is true, and yet not be able to share the grounds The main idea b justification or knowledge with another party, not because of any fault on her part (epistemic or communicative), but because those grounds – even if appropriate – are relative to a subjective factor that is not shared among parties. 

Notice also that this epistemic kind of faultless debate is also closely related to subjective truth. Commonly when there is a faultless debate, one can find an epistemic proposition, i.e., a proposition about the epistemic status o the subjective judgment or belief whose truth-value is subjective in the above sense (i.e. relative). For example, one may say that whether S is warranted to hold certain belief or to assert certain proposition is a subjective matter, because the grounds S has for holding such belief or asserting such proposition are subjective. 

Now, the central question is whether subjective truth and subjective judgment are related in such a way that (A) subjective truths can be known only subjectively (if they can be known at all) and (B) objective truths can be known only objectively (if they can be known at all).  

Originally written circa 2009