1 Introduction

In a 1911 research manuscript “Das Problem der Idealität der Bedeutungen” (“The problem of the ideality of meanings”), Husserl puts forth an idea that seems to anticipate later arguments about meaning. Having introduced the notion that the meanings of demonstrative terms cannot be separated from the intentional acts in which they are expressed and the subjects of those acts, Husserl poses the following question: “But what about two people on two celestial bodies who, in their respective surrounding-appearances, which are completely alike, present ‘the same’ objects and then orient ‘the same’ statements? Has not the ‘this’ got a different meaning in each of these cases?” (Husserl, 1987, pp. 211–212)Footnote 1 The appeal to two distinct yet qualitatively identical planets where two people utter the same expression bears a striking resemblance to Putnam’s famous Twin Earth thought experiment. In Putnam’s version, we are asked to imagine a Twin Earth, a nearly exact replica of Earth, which also includes a twin for every inhabitant of Earth. The two planets are “nearly” identical because there is one subtle but crucial difference between the two seemingly identical planets: where the molecular structure of water on Earth is H2O, it is another chemical formula XYZ on Twin Earth. Due to this difference, Putnam argues, the meaning of the term “water” is different when I use the word “water” on Earth and when my twin uses the word on Twin Earth. On Earth, the word refers to H2O; on Twin Earth, it refers to XYZ – undeterred by the fact that neither my twin nor I have any knowledge of such molecular structures. (Putnam, 1973, pp. 700–703; 1975, pp. 139–144.) Despite differences between Putnam’s argument and Husserl’s claim, both seem to appeal to the idea that external elements can determine meaning even if those elements remained indistinguishable to the subject. This idea is called externalism.

The fact that Husserl invokes some version of the Twin Earth thought experiment probably strikes many as odd. At first glance, Husserl’s phenomenology, which could be characterized as a metaphysically neutral study of consciousness or experienceFootnote 2 from the first-person perspective, seems to have much more in common with internalism rather than externalism. In fact, in The Oxford Dictionary of Philosophy, for just one example, it is written that what is distinctive about Husserl’s phenomenology “is that the content is immanent, existing within the mental act, and anything external drops out as secondary or irrelevant to the intrinsic nature of the mental state” (Blackburn, 2016). While this quote is emblematic of somewhat widespread internalist ways of understanding Husserl’s phenomenology, especially among non-Husserlian philosophers, many contemporary Husserl scholars have deemed the internalist interpretation of Husserl misguided. However, these scholars, who reject the internalist interpretation, disagree whether the correct way to read Husserl is as an externalist (Beyer, 2000, 2013; Smith, 2008; Crowell, 2008) or as a philosopher whose thought escapes the internalism-externalism framework altogether (Murchadha, 2003, 2008; Zahavi, 2004, 2008, 2017; Alweiss, 2009; Man-To, 2014). In this paper, I take Husserl’s “Twin Earth” passage as a point of departure and propose an externalist interpretation of Husserl supported by additional textual material as well as a broader realist understanding of phenomenology.

In the following Sect. 2, I define internalism and externalism. I analyze Putnam’s argument for externalism into two distinct arguments, the other people argument and the real world argument, and I then use this distinction to make a further distinction between two forms of externalism: weak externalism, which only accepts the other people argument, and strong externalism, which accepts both arguments. In Sect. 3, I analyze Husserl’s 1911 manuscript and assess various readings of it. Although the manuscript remains open to both internalist and externalist readings, I maintain that its “Twin Earth” passage is best understood as indicative of externalist ideas that Husserl presents in his other work. Before moving to this additional evidence, I motivate the externalist interpretation by criticizing the internalist interpretation and defending a realist reading of Husserl in Sect. 4.1. In the final Sect. 4.2, I claim that existing externalist interpretations are either weak or ambivalent with regard to the real world argument. Conversely, I argue for a strong externalist interpretation. I do this by evaluating Crowell’s excellent but somewhat embryonic externalist interpretation. Although the starting point of this paper is historical, the proper development of an externalist Husserl interpretation also requires clarifying what has become a rather convoluted conversation in the literature. To bring clarity to the internalism-externalism debate and the Husserlian contributions to the debate, it is necessary to tease out the nature of internalism and externalism as well as Husserl’s transcendental phenomenology. Therefore, this study also serves a systematic purpose.

2 Weak and strong externalism

Putnam defines semantic externalism against the background of what he calls “the traditional view” of meaning which has come to be known as semantic internalism. Putnam identifies two key assumptions of the view: (1) “knowing the meaning of a term is just a matter of being in a certain psychological state” and (2) “the meaning of a term (in the sense of ‘intension’) determines its extension (in the sense that sameness of intension entails sameness of extension)” (Putnam, 1975, pp. 135–136; see also Putnam, 1973, p. 700). The second assumption is well known in philosophy of language: if two terms share the same meaning, then they share the same reference, but even if two terms differed in their meanings, they could still share the same reference (e.g., “the Morning Star” and “the Evening Star” both refer to the planet Venus). The first assumption of the traditional view, however, requires some further clarification. By “certain psychological state,” Putnam does not mean any psychological state whatsoever but precisely a narrow mental state in the more technical sense of methodological solipsism. Such mental states are those that do not presuppose “the existence of any individual other than the subject to whom that state is ascribed” (Putnam, 1975, p. 136). The first assumption thus entails that knowing the meaning of a term does not require knowing anything beyond the subject’s narrow mental states, which indicates that meaning depends on nothing but narrow mental states. The first assumption is then the claim that narrow mental states alone determine meaning (see Putnam, 1975, p. 139). If the narrow mental states of two subjects were the same, then their meanings would be the same as well. One might therefore summarize semantic internalism in two claims: (1) narrow mental states alone determine meaning and (2) meaning alone determines reference.

In contrast to internalism, externalism can be defined as the denial of the first claim: not only internal narrow mental states but also external factors, which do require the existence of something else than the subject, determine meaning. Although Putnam focuses on arguing against the first claim, his arguments also modify the second claim. If meaning is not determined by narrow mental states alone but also by external factors, then the second claim loses one of its important implications. This implication is that two terms can differ in their intension when sharing their extension but not vice versa: two terms could not differ in their extension while sharing their intension. However, once the first claim of internalism is rejected, the second claim no longer has this restricted implication. On the contrary, Putnam maintains, two terms could differ in their extension even if they had the same intension. Then, the second claim becomes either the claim that meaning alone determines reference but meaning can be externally determined or the claim that meaning does not alone determine reference, but it co-determines reference with other contextual factors.

Putnam presents two arguments against the first claim of internalism. The first argument is the Twin Earth argument: If we imagine two people on two seemingly identical planets, which are different only in the molecular structure of water, and the two people utter the sentence “water is wet,” we are inclined to say, Putnam argues, that the meanings of their sentences are different. According to Putnam, the meanings are different due to the difference in the hidden structures of the natural kinds of water on Earth and water on Twin Earth. Putnam goes as far as claiming that even if we imagined my twin and I using the word “water” in the year 1750, which predates modern chemistry, the meanings of our words would still differ because of these hidden structures (Putnam, 1975, p. 141). The second argument against the first claim of internalism is the division of linguistic labor. The idea is that even if I was unable to distinguish certain natural kinds from one another, there might be people in my linguistic community who could. If I was unable to distinguish between, say, two kinds of trees, elms and beeches, the meanings of the terms “elm” and “beech” would nevertheless differ even in my usage of the terms insofar as there were people in my linguistic community who could make the distinction (Putnam, 1975, pp. 143–144). “Thus the way of recognizing possessed by these ‘expert’ speakers,” Putnam writes, is “possessed by the collective linguistic body, even though it is not possessed by each individual member of the body, and in this way the most rechercé fact about water may become part of the social meaning of the word while being unknown to almost all speakers who acquire the word” (Putnam,1975, p. 145).

Both arguments reach the same conclusion that “the psychological state of the individual speaker does not [alone]Footnote 3 determine ‘what he means’” (Putnam, 1975, p. 192). Meaning can also depend on the linguistic community and its conceptions as well as the real world and its hidden structures. Although these arguments are sometimes discussed together, Putnam distinguishes them, describing “semantic externalism and the division of linguistic labor” as his “two doctrines” (Putnam, 1996, p. xxi). “[T]he Twin Earth thought experiment,” Putnam writes, “was only one of a number of arguments that I used to argue that, contrary to these traditional views, knowledge of meanings is not something that is possible for a thinker in isolation, and that it presupposes both interactions with the world and interactions with other language users” (Putnam, 1996, p. xvi). To Putnam, the traditional view of meaning has “the tendency to treat cognition as a purely individual matter and the tendency to ignore the world, insofar as it consists of more than the individual’s observations. (…) Traditional philosophy of language, like much traditional philosophy, leaves out other people and the world; a better philosophy and a better science of language must encompass both.” (Putnam, 1975, p. 193; see also Putnam, 2013, p. 197.)

Putnam’s two arguments concern language and natural kind terms specifically, but further research has extended the scope of externalism to other “non-natural” terms (e.g., terms about ordinary artificial objects such as pencils and phones) as well as mental states (see e.g., McGinn, 1977, 1989; Burge, 1979, 1982). The reasoning is quite straightforward because linguistic expressions typically express mental states. “What the twins say is what they mean,” Gabriel Segal explains, “i.e., the content of the belief they express is just the content of the sentence they utter,” due to which “what the twins believe is different too” (Segal, 2000, p. 24). For this extended version of externalism, Burge devises a variation of the Twin Earth argument: “[W]e can imagine Alfred’s believing of apple 1 that it is wholesome, and holding a true belief. Without altering Alfred’s dispositions, subjective experiences, and so forth, we can imagine having substituted an identically appearing but internally rotten apple 2. In such a case, Alfred’s belief differs while his behavioural dispositions, inner causal states and qualitative experiences remain constant.“ (Burge, 1982, p. 97) Although Putnam did not make this extension himself, he has subsequently endorsed it (Putnam, 1996, p. xxi).

In this paper, I discuss externalism in the extended sense as a thesis about both linguistic and mental contentFootnote 4. Despite this broader scope of externalism, the above analysis of Putnam’s two arguments can still be applied. In order to emphasize the departure from mere linguistic content, I re-name these arguments the real world argument (Twin Earth) and the other people argument (division of labor), following the two blind spots of “traditional philosophy” that Putnam uncovers. Where the real world argument claims that the content of a subject’s mental states and linguistic expressions can depend on the real world and its hidden structures, the other people argument claims that content can depend on other people and their conceptions. Such conceptions no longer indicate merely expert beliefs about natural kinds and their correct terminology but also beliefs and ideas about the world and its objects more broadly (not restricted to natural kinds and language) in the community of subjects. Based on this distinction between the two arguments, I distinguish between two kinds of externalism: (1) weak externalism that accepts the other people argument but rejects the real world argument and (2) strong externalism that accepts both arguments. This distinction is useful because the arguments concern different philosophical questions, and there are clearly disparate reactions to them in the literature. Although all internalists reject the real world argument, some have tried to integrate the division of linguistic labor, or the other people argument, in their internalist accounts of meaning (e.g., Zemach, 1976, pp. 124–127; Mellor, 1977, p. 304). In addition, McGinn, who is an externalist, defines internalism, unlike Putnam and Burge, as allowing “that what one believes depends upon facts about one’s community” (McGinn, 1989, p. 2n5). Although it has been argued that it is not possible to combine the division of linguistic labor (i.e., the other people argument, if construed more broadly) with internalism (see Sterelny, 1983; see also Pessin & Goldberg, 1996, p. 59n1), the readiness of some philosophers to incorporate the argument into internalism reflects a difference between the real world argument and the other people argument. Even if internalists were unable to subscribe to the other people argument, it would nevertheless make sense to distinguish between two versions of externalism, which differ in terms of these arguments.

In this distinction, weakness of externalism stems from denying that things in the real world could determine content, whereas strength comes from affirming their ability to do so. Although weak externalism is externalism in rejecting the idea that only narrow mental states determine content, strength is important for the philosophical motivation of externalism. If one just wanted to reject the internalist claim that the individual’s narrow mental states alone determined meaning, it would be sufficient to just employ the other people argument. The reason why many externalists nevertheless use the real world argument as well seems to be that if externalism only broadened the concept of meaning socially (to the “collective body” of other people), externalism could not secure the identity of meaning irrespective of changes in the conceptions of the community. Yet this seems to be precisely something externalism often tries to achieve. For one, Putnam claims that the scientific term “atom” has retained the same meaning despite its historically changing definitions because its identity is secured by hidden physical structures that partly determine its meaning. “[W]hen physicists gave up the idea that atoms are little solar systems (the Bohr model),” Putnam writes, “they did not change the reference of the term ‘atom’; rather, they arrived at a better idea of what atoms are” (Putnam, 2013, p. 194). Only the real world argument is able to introduce a dependence relation between meaning and reality, which gives rise to what McGinn calls “the philosophical significance of externalism” (McGinn, 1989, p. 9). If mental states depend for the identity of their contents on the real world, then externalism presupposes realism about the real world (i.e., that the real world exists independently of the subject or subjects), implying that externalism is incompatible with metaphysical idealism (according to which the existence of the real world depends on the subject or subjects). “[I]f we have reason to believe externalism (about some class of mental states),” McGinn writes, “then we thereby have reason to favour realism about the subject matter of those states” (McGinn, 1989, p. 11). Although internalism is not in any conflict with realism, internalism remains neutral with regard to both realism and idealism. In contrast, externalism precludes idealism and entails realismFootnote 5.

In talking about internalism and externalism, one inevitably ends up using the notions of internality and externality. To Putnam, the line between the internal and the external seems to be the body of the individual subject. If something occurs within the subject’s body, then it is internal, and vice versa. Although this way of drawing the line between internality and externality has received some criticism (see e.g., Jacquette, 2013, pp. 69–70; Wikforss, 2008, p. 162; Gertler, 2012), I do not intend to challenge the widely accepted “spatial criterion” here. However, I will use a different criterion introduced by Farkas (2003) because it better highlights the difference between internalism and externalism that is most significant when it comes to approaching the debate in relation to Husserl. In brief, according to Farkas’ criterion, internalism restricts the determination of content to factors that are subjectively distinguishable, whereas externalism also allows factors that are subjectively indistinguishable to determine content (see Farkas, 2003, pp. 196–197; see also Farkas, 2008, pp. 82–84). To lend Burge’s example mentioned above, the internal rottenness of the second apple is a subjectively indistinguishable feature in my perceptual experience of the second apple because I cannot distinguish the second apple from the first. Based on Farkas’ criterion, I define the positions in the following way: internalism claims that content is solely determined by factors that are subjectively distinguishable, whereas externalism claims that content can also be determined by factors that are subjectively indistinguishable. The difference between weak and strong versions of externalism then concerns the question whether the subjectively indistinguishable factors must at least be distinguishable by some members of the community (weak externalism) or not (strong externalism). Although this alters the formulation of externalism, the definition maintains the philosophical significance of (strong) externalism. If subjectively indistinguishable features of things in the real world can determine content, as strong externalism maintains, then it becomes impossible to make the idealist claim that the existence of the real world depends on the subject or subjects. The subjectively indistinguishable factors are indicative of a real world out there beyond the subject’s epistemic reach, and the claim that they can determine content thus entails realism.

Even if this was not the best way to characterize the difference between internalists and externalists in contemporary philosophy, as Farkas proposes, it would be, I believe, the most useful way to approach the issue when it comes to discussing the debate in relation to Husserl. This is because Husserl’s phenomenology, as Zahavi correctly points out, “plays havoc” with “the inner/outer division” that is so pertinent to the debate (Zahavi, 2017, p. 118). For example, Husserl writes that “the transcendental ego has no exterior; that is completely senseless” (Husserl, 2003, p. 179) and calls the idea of an outside beyond transcendental subjectivity “nonsense” (Husserl, 1973b, p. 117/84). Husserl also challenges the separation between “internal” experience of one’s own experiences and “external” experience of physical things because all experience is essentially subjective and “part of an indivisible totality” (Husserl, 1976b, p. 224/220) and talks about a “new ‘internal world’” that his phenomenology uncovers (Husserl, 1976b, p. 246/243). “What is a mutual externality from the point of view of naïve positivity or objectivity is,” Husserl writes, “when seen from the inside, an intentional mutual internality” (Husserl, 1976b, p. 260/257). It therefore seems arduous to apply the spatial distinction to Husserl’s texts concerning philosophical issues around meaning. Zahavi in fact concludes that “the internalism/externalism divide loses its relevance” within “the transcendental framework of Husserl’s thinking” (Zahavi, 2017, p. 119) and recommends that we reject the internal-external divide, which “might at least permit us to avoid letting our investigation be guided by misleading metaphors” (Zahavi, 2017, p. 119). While I think Zahavi is correct in pointing out this incongruity between Husserl’s transcendental phenomenology and the internalism-externalism framework, I think the issue does not arise if the internal-external divide is not understood spatially but epistemically, that is, as a divide between subjective distinguishability and indistinguishability. Even if Husserl “played havoc” with the traditional spatial divide, his descriptions of experience would nevertheless have to deal with the line between what is subjectively distinguishable and what is not. Although some of Husserl’s texts contain phrasings that reject the internal-external divide in one sense, I do not think this fact invalidates the significance of the internalism-externalism framework and its application to Husserl’s phenomenology altogether. On the contrary, I think it invigorates the philosophical questions at the heart of the debate. It is precisely because of these difficulties with applying the internalism-externalism framework to Husserl’s phenomenology that the investigation is not only historically but also systematically significant. The historical study of whether Husserl might be an internalist or an externalist calls for a re-evaluation of central issues regarding Husserl’s transcendental phenomenology, the philosophical significance of externalism, and the nature of the content-reality relation that externalism affirms. I will return to these issues in Sect. 4.

3 Internalism and Husserl’s Twin Earth

I will start the investigation of Husserl’s relation to the internalism-externalism debate by looking at Husserl’s employment of a hypothetical scenario akin to Putnam’s Twin Earth in the 1911 manuscript. It is first important to grasp the context in which Husserl presents the scenario before dealing with it per se. The 1911 manuscript studies the problem of the ideality of meaning. Ideality designates the idea that “meanings should be ‘ideal’-identical against the flow of stating, expressing, naming, and the associated act of meaning, so they should not be bound to the singular acts and thus not to the meaning-giving subjects.” (Husserl, 1987, p. 202) When I think about the fact that water consists of oxygen and hydrogen, and when I later make a judgment about it, the meaning of these two separate acts remains the same. The ideality of meaning indicates its stable identity in “the flow” of acts. Although this conception of meaning works well for acts whose meanings are invariant, there seem to be singular acts whose meanings are not. The prime examples of such singular acts are what Husserl often calls “occasional judgments” (i.e., judgments involving demonstratives or indexicals, that is, context-dependent terms), but also perceptual experiences that are inevitably demonstrative. This is the problem of the ideality of meaning.

The 1911 manuscript can be seen as correcting Husserl’s earlier treatment of the problem in Logische Untersuchungen [1900–1901] which Husserl, already in the 1913 second edition of that book, deems “an act of violence” (Husserl, 1975, p. 13/7). In Logische Untersuchungen, Husserl claims that occasional judgments can be translated into objective judgments. “The content meant by the subjective expression, with sense oriented to the occasion,” Husserl writes, “is an ideal unit of meaning in precisely the same sense as the content of a fixed expression. This is shown by the fact that, ideally speaking, each subjective expression is replaceable by an objective expression which will preserve the identity of each momentary meaning-intention.” (Husserl, 1984, p. 95/223.) Husserl’s idea seems to stem from his attempt to shift the problem from meanings to acts: “Rightly seen, however, such change in meanings is really change in the act of meaning. (…) But the meanings themselves do not alter.” (Husserl, 1984, p. 96/224) Thus Husserl’s early treatment fails to appreciate what has later become known as “the problem of the essential indexical” (Perry, 1979): the impossibility of providing “objective” translations for essentially indexical expressions. However, fragments of an alternative treatment of the problem of the ideality of meaning can be found in the 1911 manuscript.

In addition to abandoning the idea of translation, the 1911 manuscript also addresses the problem as a broader issue not limited to philosophy of language but concerning singular acts overall (see also Husserl, 1975, p. 13/7). Hence new and wider terms “empirical meaning” and “empirical presentation” take over “occasional meaning” and “occasional judgment” as denoting the meanings and presentations of singular acts. Husserl defines empirical meaning by distinguishing it from pure meaning: where pure meanings are the meanings of acts that are directed toward something ideal (e.g., mathematical objects), empirical meanings are the meanings of acts that are directed toward something “real” and “individual” (Husserl, 1987, p. 202). The problem of the ideality of meaning concerns empirical meaning because the ideality of pure meaning is obvious: the content of different acts of thought about the same number, for example, is quite separable from the acts of thought themselves. Conversely, the ideality of empirical meaning seems questionable. Since the ideality of meaning indicates the identity of meaning, the 1911 manuscript unfolds as an inquiry into the conditions of identity for empirical meaning. “When can individual thing-meanings, empirical meanings in general (which orient themselves according to different acts of empirical presentations),” Husserl asks, “be identical?” “Or rather,” Husserl continues, “when can meaning intentions, based on different empirical presentations, carry same meanings?” (Husserl, 1987, p. 206).

Initially, Husserl seems to provide an internalist response: “Now I must be able, one will say, to recognize the object presented the second time as the same as I presented it the first time” (Husserl, 1987, p. 206). This seems to commit Husserl to internalism. If I must be able to recognize the object as identical in different acts for the empirical meaning to be identical, then the identity of empirical meaning cannot depend on further factors that are subjectively indistinguishable. However, there is further textual evidence indicating that Husserl thought otherwise, which I will discuss in Sect. 4.2. Yet, even without such evidence, it is possible that Husserl is here, using the expression “one will say” (“wird man sagen”), outlining what he takes to be the typical initial response. I think this is in fact the case and that the ensuing Twin Earth thought experiment is used to challenge the idea. Having presented the ideas that empirical meaning is “inseparable” from its act and subject and that the subjective distinguishability of the object’s identity is necessary for the identity of meaning, Husserl introduces a challenge to these very ideas. “But what about two people on two celestial bodies,” Husserl inquires, “who, in their respective surrounding-appearances, which are completely alike, present ‘the same’ objects and then orient ‘the same’ statements? Has not the ‘this’ got a different meaning in each of these cases?” (Husserl, 1987, pp. 211–212.)

Where the preceding statements (that the meanings of demonstrative expressions cannot be separated from the acts in which they are expressed, and that the subjective recognizability of the identity of the object is necessary for the identity of the meaning of the act) seem to be challenging the idea that meaning is always ideal, this question seems to challenge these statements. Rather than answering the question directly, Husserl answers it by discussing a modification to the thought experiment in the subsequent sentence: “Let us change the example a bit. Let there be two similar environments created on Earth and two people in the same position towards them, both having completely the same appearances, expressing themselves in the same words, etc. Do the words have the same meanings on both sides?” (Husserl, 1987, p. 212.) “If we say ‘no,’” Husserl continues, “we do so based on the knowledge of the presupposed state of affairs.” This knowledge stems from us describing these two experiences in “a unity of consciousness” in which we arrive at the point that “we no longer have two completely similar experiences.” Rather these experiences become intertwined with “conceptions, empirical positions, [and] components of experience that relate to the experiential context.” To Husserl, this exercise points out that the experiences have not been “exactly the same for both people in the same surroundings in the first place.” “They would have been the same,” Husserl writes, “if we pretended that the area of sameness went so far that two completely similar people had grown up in a completely similar world of experience (namely a completely similar area of experience related to them), as above with the celestial bodies and even further.” (Husserl, 1987, p. 212.) What Husserl ends up implying here is that we precisely do pretend that the experiences were the same for both people in the thought experiment of the two seemingly identical celestial bodies (“as above with the celestial bodies”). If we do pretend, as we often do in thought experimentsFootnote 6, then the answer to the question is that the empirical meanings of the two utterances of “this” differ.

The important question then is why they differ. Here things do get a little unclear in Husserl’s manuscript. Rather than providing us with an explanation about the difference in meaning, Husserl switches between rejecting certain alternatives and vaguely pointing toward others. Based on the idea that two experiences are never completely alike, Husserl ponders whether this definitively shows that “the identity of empirical meaning depends on the empirical act of presenting and thus also the empirical act of meaning” (Husserl, 1987, p. 212). However, as Husserl immediately points out, “the meaning does not depend on the individual act,” nor does it depend “on the fact that the empirical intuitions, which are brought to unity, really belong to the same objectivities” (Husserl, 1987, p. 212).

It is this claim that warrants internalist interpretations of the manuscript. Beyer, who was the first scholar to discuss Husserl’s “Twin Earth” passage, presents just such an interpretation in his first treatment of the passage (Beyer, 1996), though Beyer has changed his mind over the years toward a more externalist interpretation of the passage (see Beyer, 2000, 2013, 2017, pp. 221–222)Footnote 7. Although Beyer takes Husserl to answer the Twin Earth question affirmatively (that the meaning of “this” differs in the two cases), he initially provides an internalist interpretation of Husserl’s response. “According to Husserl,” Beyer writes, “it plays no role for the identity of direct or own-meaning [Eigenbedeutung] whether the ‘identifying consciousness is right or wrong” (Beyer, 1996, p. 179). “It is thus according to Husserl possible (and often actual),” Beyer continues, “that one can perceptually confuse a real object with another object and nevertheless realize the same empirical meaning on both sides” (Beyer, 1996, pp. 179–180). In Beyer’s internalist interpretation, Husserl takes the meaning of two acts to be identical if the object appears as the same regardless of possible indistinguishable differences between the two objects. This internalist interpretation can be combined with an affirmative answer to the initial Twin Earth question because the difference of the two utterances of “this” is due to a difference in the two subjects: they are two numerically different subjects with different experiences (Beyer, 1996, p. 183).

Following Beyer’s internalist interpretation, Joumier also reads Husserl as an internalist. Based on Husserl’s remark quoted above, Joumier rejects the externalist idea that the difference between two empirical meanings of two different empirical acts of empirical presentations was due to a difference in the real world. “The identity of meaning is in a sense,” Joumier sums up, “indifferent to the identity of the object” (Joumier, 1999, p. 57). If I have two different perceptual acts of two distinct objects, the meanings of these two acts differ only if “I believe they are different” (Joumier, 1999, p. 57). According to Joumier’s interpretation, subjectively indistinguishable features do not determine the identity of empirical meanings; their identity is rather resolved by “consciousness of unity” that unifies different presentations of the posited object into the same presentation irrespective of how “unfounded” this identification is (Joumier, 1999, p. 57). In a forthcoming publication of his on the 1911 manuscript, Dagfinn Føllesdal seems to agree with both Joumier’s and Beyer’s internalist interpretations. “The reason why Husserl called himself an idealist, is I think,” Føllesdal claims, “mainly that he was very much an internalist.” To Føllesdal, “Husserl’s main point, in this manuscript and other places where he discusses these issues, is the following: Our use of language, including singular terms, like ‘this’, is based on our present conception of the world. (…) In their normal use our words do not relate to a world of ‘the final theory’ or the world ‘as it really is,’ whatever that may mean. We refer to the world as we presently conceive of it.” (Føllesdal, forthcoming.)

These internalist interpretations find their strongest textual evidence in the passage where Husserl seems to reject the idea that the identity of empirical meaning depends “on the fact that the empirical intuitions, which are brought to unity, really belong to the same objectivities” (Husserl, 1987, p. 212). However, it is important to look at the context of this passage as Husserl continues the claim by discussing his Twin Earth thought experiment. “If I were brought from one environment to another in exactly the same way,” Husserl writes, “with the empirical intuitions being either absolutely the same or fitting together in the sense of just intuitively the same environment, and to have returned to the same point of departure, confused by the path that did not give enough clues for correctly estimating the direction of movement, the same objects count as the same for me, and then I will also say that the meanings are the same.” (Husserl, 1987, p. 212) Husserl says here that if I were in two seemingly identical environments, the objects would “count as the same for me” and therefore “I will also say that the meanings are the same.” These statements are compatible with externalism: if I was transported from Earth to Twin Earth, I would not be able to distinguish the two distinct liquids called “water,” and therefore I would certainly say (though I would be wrong) that the meanings of my two expressions are the same. Having made this point, Husserl then writes that “[t]he opinion [Meinung] of ‘this house,’ ‘this table’ etc. is then the same on both sides” (Husserl, 1987, p. 212). The German word “Meinung” is usually translated as “opinion,” which conveys the idea quite well: when the two subjects are not in the position to distinguish their objects, then their acts share something from the subjects’ perspectives, and that something can be called “opinion,” “narrow content,” or “subjective meaning.” A few sentences later Husserl seems to expand on this: “All empirical presentations, which belong to the unity of an object meant as identical (…) ground the possibility of an identical act of own-meaning” (Husserl, 1987, p. 213). Husserl then claims that if something “counts as the same (…), it is meant as one, and if that is really so, then the own-meaning [Eigenbedeutung] is the same” (Husserl, 1987, p. 213). The way I see it, the concept of own-meaning is equivalent with the concept of opinion. It denotes the aspect of meaning that is solely determined by subjectively distinguishable factors. If “I mean a and I identify a with b, then I also identify the meaning of a with that of b, namely as own-meaning. If I afterwards distinguish between a and b, then now the ‘this’ means different for me here and there.” (Husserl, 1987, p. 213.)

However, in addition to opinion or own-meaning, Husserl seems to have other concepts of meaning. In fact, Husserl is quite explicit about it. “We must distinguish,” Husserl writes, “the valid relation of meaning to the object and the meaning per se, relating precisely as opinion to the meant object” (Husserl, 1987, p. 213). Husserl seems to suggest a distinction between the level of meaning that can have a valid relation to the real object and the level of meaning as opinion that cannot as it merely relates to the meant object. I speak of levels of meaning because rather than suggesting a distinction between distinct types of meanings (as so-called two-component theories of narrow and broad content sometimes do), Husserl is, it seems to me, distinguishing between layers of meaning that can belong to the same act.

The distinction between these two levels, own-meaning and meaning that can relate to the object, helps explain some of Husserl’s seemingly internalist claims in the manuscript as claims about the former, but it also facilitates the understanding of the “Twin Earth” passage in the manuscript. In a word, the distinction makes it easier to articulate how the passage is read: that is, whether one takes the passage to be concerned with own-meaning or meaning that can have a valid relation to the object. In their internalist readings, Beyer and Joumier take the passage to be concerned with own-meaning. Therefore, to them, the difference of meaning is due to the difference in the subjects. However, it is worth noticing that this makes Husserl’s question trivial. It is now just the question whether the meanings of two demonstrative utterances made by two different people in two distinct places differ. This question is trivial because demonstrative utterances are always context-dependent, and if the two people are in two distinct contexts, then the meanings of their utterances are obviously different.

Putnam is not concerned with this trivial question. The innovative nature of Putnam’s argument is precisely the suggestion that even non-indexical expressions, such as those involving natural kind terms, might exemplify similar context-dependence – they might carry “an unnoticed indexical component” (Putnam, 1975, p. 152). Although Husserl’s “Twin Earth” passage does not exemplify the same innovation, I do not find Husserl’s question trivial. It is possible to read Husserl’s question as concerned with meaning that can have a valid relation to the object rather than own-meaning (i.e., opinion). Perhaps the best argument for reading the passage in this way is that it makes more sense for this reading why Husserl introduces the seemingly identical celestial bodies in the first place. If the celestial bodies were relevant at all, then it would make sense that Husserl is inquiring about a level of meaning that can have a relation to them or their properties. If Husserl were just inquiring about own-meaning and intending to make the trivial claim that own-meaning is different for distinct people, then it would make little sense to even introduce the celestial bodies. They would serve no purpose for the trivial argument. If one reads the passage as concerned with meaning that can have a valid relation to the object, then the question is no longer trivial. If the passage is read so, then it seems that Husserl is committed to externalism.

In summary, my reading of Husserl’s “Twin Earth” passage is that while Husserl presents some internalist ideas in the 1911 manuscript, he challenges them by employing the “Twin Earth” example. However, given that it is by no means self-evident that Husserl abandons all these internalist ideas in this research manuscript, the reading remains tentative and the manuscript open to opposing readings. Even if one did not accept my reading, I think the “Twin Earth” passage, at the very least, shows that Husserl has interest in externalist ideas. I go through further textual evidence to uncover these ideas in Sect. 4.2.

4 Externalism and the sources of the identity of meaning

4.1 Motivating the externalist interpretation

Before turning to this further textual evidence, it is important to motivate the externalist interpretation. One of the main motivations for pursuing an externalist interpretation of Husserl is the shortcoming of the internalist interpretation, which is that the internalist interpretation falsely commits Husserl to methodological solipsism. Methodological solipsism is the idea that mental states do not depend on anything except the individual whose mental states they are. The implication is that a subject’s mental state can be individuated by appealing solely to that mental state, its properties, and its relations to other mental states of the subject (Fodor, 1980). Joumier is explicit about this in discussing “the methodological solipsism” of Husserl’s transcendental reduction (Joumier, 1999, p. 62), but Beyer’s (1996) early internalist interpretation also seems to be committed to it since internalism entails methodological solipsism.

Internalism and methodological solipsism have often been applied to Husserl based on readings of his phenomenological methodology. For example, Dreyfus and Hall write: “The job of the phenomenological psychologist is to examine the activity that makes reference possible while remaining uncommitted as to whether, in any given case, or even in general, reference is in fact achieved. This abstention, which Husserl calls the phenomenological epoche or the bracketing of existence, has, he insists, important methodological implications for psychology.“ (Dreyfus & Hall, 1984, p. 14.) “[T]his bracketing of the concerns of naturalism, along with the implicit denial of the relevance of the causal component of reference,” they write, “makes Husserl a methodological solipsist” (Dreyfus & Hall, 1984, p. 15). Reflecting upon this kind of understanding of Husserl, Cerbone (2020) writes that Husserl’s descriptions of the epoché as “bracketing, parenthesizing, or putting out of play” seem to exemplify “Husserl’s insistence on purifying consciousness, where this means isolating or insulating consciousness as much as possible from any considerations concerning goings-on in the ‘external’ world.” Therefore, Cerbone writes, Husserl’s phenomenological methodologyalso seems to yield “a number of other methodological and theoretical commitments, chief among them methodological solipsism, anti-naturalism, metaphysical neutrality, and a resolute internalism about mental content.“ (Cerbone, 2020, p. 605.)      

These readings tend to appeal to the methodology of Husserl’s transcendental phenomenology, namely the epoché, and, therefore, one must look more closely at the epoché in order to evaluate these readings. The epoché is a methodological step that Husserl uses to bracket a general implicit assumption about the real world. In Ideen I [1913], Husserl describes the epoché as “a certain withholding of judgment” (Husserl, 1976a, p. 64/54) that puts “out of action the general thesis belonging to the essence of the natural attitude” (Husserl, 1976a, p. 65/55). The natural attitude is the attitude in which I live, experience things, encounter others, and conduct scientific experiments. The general thesis of the natural attitude is the implicit assumption that everything I experience in the world exists independently of my experience (Husserl, 1976a, p. 62/53). Husserl writes that “I find the ‘reality’ (the word already says as much) to be there in advance and I also take [it] as it affords itself to me, as being there. No doubt or rejection of anything given in the natural world changes anything in the natural attitude’s general thesis.” (Husserl, 1976a, p. 61/52, translation modified.) Since the general thesis is the implicit assumption that reality is “there in advance,” one can think of the general thesis as naïve or common-sense realism, i.e., the view that objects of reality exist independently of one’s experience of them. The epoché is the bracketing of this implicit assumption; it is the suspension of judgment with regard to its truth-value.

However, this bracketing of the general thesis does not mean that I denied or even doubted the natural attitude (see Husserl, 1976a, p. 65/55–56). In fact, Husserl could not reject the natural attitude because the purpose of transcendental phenomenology is to study the natural way of experiencing the world. In a 1934 letter to Baudin, Husserl even claims that “[t]he method of the phenomenological epoché and reduction presupposes the existence of the world” (Husserl, 1994, p. 16). In his 1930 “Nachwort” to Ideen I, also published in 1932 as a preface to the book’s first English translation, Husserl writes that the existence of the real world “is quite indubitable” and that the “sole task” of phenomenology “is to clarify the meaning of this world, the precise sense in which everyone accepts it, and with undeniable right, as really existing” (Husserl, 1930, p. 562/22). The epoché is the method which Husserl uses to begin studying this meaning of the world.

This overview of the epoché reveals two problems in the methodologically solipsist reading. First, since methodological solipsism denies that mental states could depend on anything except the individual whose mental states they are, the epoché is misunderstood as bracketing all concerns with extra-individual things. Given the characterization of phenomenology as studying the meaning of the world, however, it seems that Husserl does not aim at closing off extra-individual things from phenomenology. On the contrary, Husserl writes explicitly that “the transcendental epoché is not to be misunderstood in the respect that the being and the being such-and-such of the world should remain out of question” (Husserl, 1959, p. 465). Having performed the epoché, “I have lost nothing that was there for me in the state of naïveté, and in particular nothing that showed itself to me as existing reality” (Husserl, 1989, pp. 174–175/495). “Precisely in this way,” Husserl continues to say, “I have brought into play a new dimension of questions never asked before and precisely about this existent reality: Only through the answering of these questions can concrete, full being and the definitive, complete truth come to light about this world.” (ibid.) Second, the methodologically solipsist reading characterizes the epoché as turning the reflective gaze away from objects in the world to a realm of mental representations, but this is incorrect. The epoché is not a reversal of reflection but rather a larger shift of attitude irrespective of what is being reflected upon. “The renunciation of the world, the ‘bracketing of the world,’” Husserl specifies, “did not mean that henceforth the world was no longer our focus at all, but that the world had to become our focus in a new way, at a whole level deeper” (Husserl, 1989, p. 173/494). It seems most plausible to understand Husserl as saying that the world investigated by phenomenology is no different from that investigated by the sciences or encountered in everyday life except for a shift of attitude towards the world (see also Husserl, 1976a, p. 3/3).

These descriptions of the epoché and phenomenology also seem to complicate the somewhat common understanding of Husserl as an idealist. Given Husserl’s tendency to employ language that lends itself to different kinds of idealist interpretations, Husserl has sometimes been interpreted even as a metaphysical idealist to whom the existence of the world depends on the existence of consciousness (see e.g., Philipse, 1995; Smith, 2003). Without going into a more detailed critical assessment of metaphysical idealist interpretations of Husserl (see e.g., Ameriks, 1977; Sebold, 2014; Hardy, 2014, pp. 176–185; Zahavi, 2017, pp. 37–40), I will just say that I think the main reason why some have read Husserl as a metaphysical idealist is Husserl’s concept of constitution. Contrary to idealist presumptions, however, it seems that Husserl does not use the concept to suggest that consciousness constructs reality, but rather that in virtue of its features, consciousness conditions the manifestation of reality, enabling one to experience the world as real (see e.g., Zahavi, 2003, p. 73; Ströker, 1993, pp. 104–107). Husserl writes that the “production of a truly existing object does not mean that the activity brings forth the object from nothing, but that, on the contrary, just as objects are already pregiven, an objective environment is always already given to us” (Husserl, 1939, p. 33–34/37). Unlike some idealist interpretations maintain, I think the concept of constitution thus entails a kind of realism: if experience enables the manifestation of something worldly, there needs to be a world to be disclosed.

Although I think Husserl should be understood as a kind of realist, this does not take away the fact that Husserl is also a transcendental idealist. In Cartesianische Meditationen [1929], Husserl writes that “phenomenology is eo ipso ‘transcendental idealism’” and that “[o]nly someone who misunderstands either the deepest sense of intentional method, or that of transcendental reduction, or perhaps both, can attempt to separate phenomenology from transcendental idealism” (Husserl, 1973b, p. 118–119/86). However, I do not think Husserl’s transcendental idealism should be identified as metaphysical idealism, as Philipse (1995) does. In the same paragraph from Cartesianische Meditationen, Husserl immediately emphasizes that phenomenology is transcendental idealism “in a fundamentally and essentially new sense” (Husserl, 1973b, p. 118/86). Husserl further explains that in his sense transcendental idealism “is sense-explication achieved by actual work,” thus again equating transcendental idealism with phenomenology (Husserl, 1973b, p. 119/86). I think this “new sense” of transcendental idealism as “sense-explication” is best understood precisely as Husserl’s phenomenology was characterized above: as the investigation of the meaning of the world that is uncovered by the epoché (i.e., how the world is experienced as real). To study this meaning, one needs to investigate the conditioning or enabling of the manifestation of the world, that is, constitution. If constitution, as the enabling of the manifestation of the world, entails a kind of realism, as I claimed it does, then Husserl’s transcendental idealism does not seem to conflict with realism. In fact, Husserl seems to make this point himself. Even when Husserl uses the term “idealism” in his “Nachwort” to Ideen I, he emphasizes that “phenomenological idealism does not deny the positive existence of the real world and of Nature”; on the contrary, “[t]hat it [the real world] exists (…) that is quite indubitable” (Husserl, 1930, p. 562/21).

There is not enough space to explore Husserl’s transcendental idealism (see e.g., Zahavi, 2017, pp. 94–108) and its connection to realism nor argue for a realist interpretation of Husserl more fully here (see Ameriks, 1977; Hardy, 2014). Suffice it to say that just as much as Husserl associates himself with idealism, he also does so with realism. For one, in the 1934 letter to Baudin, Husserl claims: “No ordinary ‘realist’ is as realistic and concrete as I, the phenomenological ‘idealist’ (a word that I by the way no longer use).” “The method of the phenomenological epoché and reduction,” Husserl continues, “presupposes the existence of the world, just as it was and is for us, and we reflecting in this method – always I, the reflecting one – are in the fully concrete world.” (Husserl, 1994, p. 16.) If Husserl is to be understood as a realist, Husserl’s realism is not any specific metaphysical position, but rather the broad common-sensical idea that the world exists independently of the existence of consciousness. “There can be no stronger realism than this,” Husserl writes, “if by this word nothing more is meant than: ‘I am certain of being a human being who lives in this world, etc., and I doubt it not in the least’” (Husserl, 1976b, pp. 190–191/187). Although there are texts in which Husserl describes himself differently and makes seemingly idealist claimsFootnote 8, the passages just mentioned seem to complicate those texts at least to the extent that Husserl most likely means something else by “idealism” and such claims than the kind of metaphysical idealism that is precluded by strong externalism. Husserl’s transcendental idealism indicates some kind of dependence on consciousness, but that dependence is not necessarily metaphysical. I will return to the nature of this dependence in the next section.

4.2 Worldly meaning and the meaningful world

Given that Husserl cannot be understood as an internalist (since that would incorrectly imply that Husserl is a methodological solipsist) and that his phenomenology can be understood within a realist framework, it becomes possible and quite reasonable to pursue an externalist interpretation of Husserl. I have already mentioned that the “Twin Earth” passage in the 1911 manuscript invokes externalist ideas in Husserl’s work, but it is important to notice that the externalist ideas, somewhat tentative in the 1911 manuscript, do not come out of nowhere. They are rather part of the development of Husserl’s thinking on the problem of meaning. In this section, I will present this broader context of Husserl’s thinking about meaning and evaluate two externalist interpretations by Smith (2008) and Crowell (2008, 2013). Finally, I outline an externalist Husserl interpretation that improves on the shortcomings of these interpretations mainly by looking at Husserl’s ideas about “worldly meaning” and “the meaning of the world” as well as the philosophical significance of externalism and McGinn’s asymmetry argument for it from a Husserlian perspective.

As said in Sect. 2, the second claim of internalism designates that two expressions with different meanings could share the same reference, but two expressions with the same meanings could not have different objects of reference. In Logische Untersuchungen, Husserl presents an opposing view. In addition to saying that “[t]wo names can differ in meaning but can name the same object,” Husserl also writes that “[i]t can happen, conversely, that two expressions have the same meaning but a different objective reference” (Husserl, 1984, p. 53/197). “The expression ‘a horse’,” Husserl writes, “has the same meaning in whatever context it occurs. (…) [I]f on one occasion we say ‘Bucephalus is a horse’, and on another ‘That cart-horse is a horse’, (…) [t]he expression ‘a horse’ employs the same meaning to present Bucephalus on one occasion and the cart-horse on the other.” (Husserl, 1984, p. 53/197–198) Husserl’s example is flawed because the word “horse” does not refer to Bucephalus and the cart-horse; the terms “Bucephalus” and “cart-horse” do. The word “horse” either refers to the natural kind of horse or indicates the horses’ shared property of being-a-horse. Despite being flawed, however, the argument shows that Husserl thinks that expressions with same meanings may have different objects, which is an externalist idea.

One of the ideas incubating in the 1911 manuscript is that there is a level of meaning that might depend on subjectively indistinguishable features. As said above, if the “Twin Earth” passage is read in internalist terms (that the difference in the meaning of “this” is due to a difference in the subjects and their acts), then the reference to the celestial bodies would be irrelevant. If the passage is read as externalist, then the passage invokes a level of meaning that contains something subjectively indistinguishable: on one level, the meaning of “this” would remain the same, as opinion or own-meaning, but, on another, it would not. Since Husserl presents a concept for this subjectively distinguishable level of meaning, since Husserl separates that notion from meaning that can have a valid relation to the object, and since Husserl uses the twin example, it seems plausible to assume that Husserl has in mind another level of meaning. Although Husserl does not elaborate on this level in the 1911 manuscript, he seems to present a similar distinction between two levels of meaning in Ding und Raum [1907] where he also reflects on the issue of the identity of meaning:

In popular terms, if I have a determinate perception of a thing, let us say of a familiar, and perhaps very familiar, thing, and if I now have a discrete second perception of the back side of this thing, then I may synthetically bring the first perception, which still lives on in memory, into a unity of identity with the new perception. In fact, the two perceptions, by their very essence, are bound together through an identity of sense. But I must immediately issue a restriction. I said, “through an identity of sense;” that would mean that the back side, as it is represented in the first perception, itself comes to appearance in the second. But is this fact given? It is indeed possible that the second perception relates to a second object, entirely similar to the first object, such that I see the first object from the front and the second from the back. Thus we would have, possibly, similarity of sense but not identity of sense, and, more generally, we would have similarity of the objective direction, i.e., direction toward something similar, but not identity as a direction toward something identical, toward one and the same object. (Husserl, 1973a, p. 154/131–132)

In Husserl’s example, I have two perceptions, a perception of the frontside and another of the backside, both of which are presumably experiences intending the same object. However, it is possible that these perceptions are not of the same object but rather of two different objects. In such a case, I could nevertheless experience the perceived front and backside as belonging to the same object. My two experiences would then lack identity of sense (Sinnesidentität) because their intended objects would not be numerically identical, but my experiences would have similarity of sense (Sinnesgleichheit) because I could not distinguish between the two objects. This distinction is of paramount importance. Since Husserl discusses the identity of sense rather than the identity of the object, he is precisely attaching the objective changes in the environment or the object to the sense or meaning. Furthermore, Husserl claims that the meaning of two experiences about two distinct objects can differ regardless of the subject’s inability to distinguish between the objects. Husserl is thus committed to the claim that there is a level of meaning that can contain subjectively indistinguishable aspects. Therefore, this passage provides counterevidence for the internalist interpretations as the identity of meaning does not require the recognition of the objects as the same.

Although Husserl thus claims that something subjectively indistinguishable can determine meaning, the source of the identity of meaning remains ambiguous. In the 1911 manuscript, Husserl at first seems to deny that the difference of meaning stemmed from a difference in reality, as both Beyer and Joumier emphasize, but Husserl is in that passage concerned with own-meaning that is precisely indifferent to reality. If one considers the “Twin Earth” passage and the concept of the identity of sense together, then Husserl seems to be introducing another level of meaning that can partly escape the subject’s epistemic reach, indicating that Husserl, like Putnam, rejects the first claim of internalism. Therefore, the source of the identity of sense is not entirely ambiguous: it does not lie solely in the subject. However, what remains unclear is whether the identity then originates from other people or the real world, i.e., whether Husserl is committed to either weak or strong externalism.

A. D. Smith’s (2008) externalist interpretation of Husserl exemplifies the response of weak externalism. Smith solves the purported conflict between externalism and Husserl’s phenomenology by proposing a new kind of externalism. Smith defines externality as “external to any individual consciousness” (Smith, 2008, p. 317). Defining externality in such a way implies that “external” things can still be internal to intersubjective consciousness or a community of subjects. Even if “the physical world may ultimately ‘reduce’ to consciousness,” Smith explains, “your consciousness cannot be reduced to mine, or mine to yours. Each individual consciousness, therefore, is ‘external’ to every other” (Smith, 2008, p. 317). Smith remarks that almost everything counts “as external to any individual consciousness, because they have an essentially inter-subjective constitution” (Smith, 2008, p. 317). Since nearly everything is constituted intersubjectively, Smith is therefore committed to claiming that practically everything is external in his sense of the term.

Although Smith rejects internalism and methodological solipsism because meaning is determined partly by other subjects, he does not appeal to the real world, which makes the externalism in his interpretation weak. One might reasonably wonder whether this is consistent. Given that other people exist in the real world, the interpretation might seem to appeal to external factors in the real world. However, as Smith himself implies, his externalist interpretation is compatible with the interpretation of Husserl as a metaphysical idealist. Smith writes that Husserl is “an out and out Idealist” (Smith, 2008, p. 316). Although Smith does not explicitly define the sense in which he uses the term “idealism” in his paper on Husserl and externalism, he refers to his other work in which he does (see Smith, 2008, p. 316n8). In this other work, Smith says that he sees Husserl as “an ‘absolute idealist’” because Husserl “subscribed to the following simple, if extreme, idealist claim: if consciousness did not exist, nothing would” (Smith, 2003, p. 179). In the same book, Smith says that “[t]here is no more to the reality of a world than the transcendental life in the community of monads unfolding in a certain harmonious way, since any real world is but the ‘intentional correlate’ of such intentionally performing life” (Smith, 2003, p. 182). “Husserl’s idealism,” Smith writes, “is the claim that physical facts and entities supervene upon consciousness,” meaning that “[p]hysical facts are ‘nothing over and above’ experiential facts – facts concerning the occurrence of actual and possible experiences in consciousness as a whole” (Smith, 2003, p. 185). Thus it seems clear that Smith is not talking about other subjects as physical things in the real world but as transcendental subjects to whom the real world can be reduced. Moreover, it is thus evident that Smith is not appealing to real things in the world but to intersubjectively constituted things that may or may not exist. It is difficult to see how this kind of extreme or absolute idealism would be compatible with any philosophically significant externalism. While this might seem to be a systematic problem rather than a historical one, it seems to me that Smith’s weak externalist Husserl interpretation bears no significant implications for Husserl interpretations either when it comes to the most important question of internalism and externalism (which is not to say that Smith’s interpretation could not have other significant implications for Husserl scholarship). This question, which both Husserl and Putnam discuss, is whether meaning could change in two acts intending seemingly identical but really distinct objects. By not addressing this question, Smith’s interpretation seems to nullify the most important internalism-externalism question for Husserl, which should be at the heart of the debate, irrespective of whether one is engaging with the question systematically or historically.

Crowell’s (2008) externalist interpretation of Husserl seems to avoid this problem, though there is some ambivalence regarding the strength and weakness of externalism in Crowell’s interpretation. According to Crowell, Husserl has been read as an internalist because Husserl does not conceive of a causal relation between content and the world. However, Crowell thinks Husserl can be read as an externalist because in Husserl’s philosophy there is a normative relation between content and the world (Crowell, 2008, p. 346). Normativity of meaning can mean many different things (see e.g., Wikforss, 2001), but Crowell applies the concept to phenomenological immanence. Where immanence is sometimes thought of as “internal to the mind,” phenomenological immanence denotes evidence or “the self-giving of something” which is contrasted with something not given “in person” but emptily intended (Crowell, 2008, p. 345). Phenomenological immanence, Crowell argues, is normatively structured because it involves the horizon. In Husserl’s phenomenology, the horizon is a structure of experience that extends experience beyond what is immediately given to the subject. The horizon consists of the currently unseen features of the object as well as other objects that are co-given alongside the object (Husserl, 1939, pp. 28–29/32–33). When I observe my laptop, I only see one side of it, but its other sides are also present in my experience. Crowell uses the term “intentional implications” to denote these features because they are implied in my experience. Intentional implications make phenomenological immanence normative. As norms, they provide a standard for success or failure. As I am experiencing, I expect things to go one way or the other, and then my experiences either satisfy or dissatisfy those expectations. For example, I expect my laptop to look a certain way if I turn it around. If I do turn my laptop around, my expectation, which is an intentional implication in the horizon of my experience, is either satisfied or dissatisfied. This is what Crowell means by the normativity of meaning.

Although Crowell does not discuss Husserl’s 1911 manuscript, he considers Putnam’s Twin Earth thought experiment from a Husserlian perspective. The gist of Crowell’s idea seems to be that even though in the year 1750 future empirical discoveries concerning molecular structure would not be available to my twin and I, they nevertheless partly determine the content of our thoughts about water since thoughts have horizons. “[B]ecause the water I perceive is not an inert datum but an element within a normative space of intentional implications (teleological interconnections),” Crowell claims, “it adumbrates avenues along which further aspects can become given – if/then structures that indicate a course of possible empirical investigation that could eventually disclose something like molecular structure. And it is these intentional links – not my causal contact with a particular substance – that determine what I am thinking about or perceiving.” (Crowell, 2008, p. 346.) Crowell concludes:

While Husserl’s reduction brackets the question of whether one can explain intentional content by appeal to causal chains and the microstructural properties of transcendent entities, he clearly denies that I could have the same content no matter what the world was like. For to have content, in Husserl’s sense, is to disclose the world in a certain way. And conversely, it is because the world is a certain way that I can disclose it thus, or ‘constitute’ it thus. If the world were different there would be different concatenations of experience, and so my mental states would also be different. (Crowell, 2008, p. 350.)

Unlike Smith, Crowell thus seems to end up affirming a dependence relation between consciousness and the world. Crowell claims that, to Husserl, experience would be different if the world was different. Therefore, Crowell claims that content is not utterly independent from the world. As said, the dependence in question is not causal but normative. This means that the dependence between content and the world goes through the intentional implications in the horizon. One possible reconstruction of the argument could be that since intentional implications depend on the world and since content depends on intentional implications, therefore content also depends on the world. People on Earth and Twin Earth would have different intentional implications and thus different contents. Since there is nothing in our current narrow mental states that could explain this difference (as they are presumed to be identical in the Twin Earth thought experiment), the difference needs to stem from somewhere else.

However, there is an ambivalence in Crowell’s interpretation when it comes to the question of weak and strong externalism because Crowell does not spell out this reconstruction himself and seems to leave the source of the identity of meaning, i.e., that “somewhere else,” open to question. Crowell’s idea that “a course of possible empirical investigation” (Crowell, 2008, p. 346) in the future could determine meaning seems to invoke the position developed by Jackman (1999) known as temporal externalism. According to temporal externalism, meaning can be determined not only by linguistic practices, which includes the past usage of terms, but also “developments in their subsequent usage” (Jackman, 1999, p. 158). “[T]he future behavior of an individual or his society,” Jackman writes, “can affect the content of his thoughts and utterances” (Jackman, 1999, p. 160). However, while Jackman’s temporal externalism makes clear that these future occurrences take place in the real world, the kind of externalism in Crowell’s interpretation seems to leave this aspect unclear. Crowell claims, in the passage quoted above, that, to Husserl, content cannot remain identical “no matter what the world was like” because “to have content” is “to disclose the world in a certain way” and the disclosure requires that “the world is a certain way” (Crowell, 2008, p. 350). Yet it is not clear how Crowell understands the term “the world.” Putnam claims that internalism ignores “the world, insofar as it consists of more than the individual’s observations” (Putnam, 1975, p. 193). I would add that weak externalism ignores the world, insofar as it consists of more than the observations and conceptions of other people. This aspect of weak externalism seems to be accurately represented by Smith’s interpretation. “There is no more to the reality of a world,” Smith writes about Husserl, “than the transcendental life in the community of monads unfolding in a certain harmonious way, since any real world is but the ‘intentional correlate’ of such intentionally performing life” (Smith, 2003, p. 182). Crowell seems to want to maintain a stronger link between the world and consciousness than Smith, but the lack of precision regarding “the world” makes it possible to read his interpretation as weak. In fact, Crowell claims that “the object can be understood phenomenologically as an open set of intentional implications” (Crowell, 2013, p. 23). If the object is, in one way or another, “a set of intentional implications,” then the dependence relation between intentional implications and the real world remains vague. Crowell criticizes internalism for failing to provide an adequate explanation for how mental representations are normative (i.e., how they have conditions of satisfaction) because internalism just appeals to the idea that normativity is “intrinsic” to mental representations (Crowell, 2008, p. 338), but I am not sure if Crowell succeeds in giving an adequate explanation either. Crowell claims that, to Husserl, normativity is “inherent” to perceptual experience (Crowell, 2013, p. 18). The problem is that if experience is normative due to its intentional implications and if normativity belongs to experience “inherently,” then the role of the real world in the determination of meaning is left ambiguous.

In the rest of this paper, rather than oppose any specific details in Crowell’s reading, I will try to resolve these shortcomings by looking at additional textual evidence, the investigation of which is also helpful in teasing out the nature of the dependence relation. Although Benoist, in his commentary on the 1911 manuscript, does not cite this additional evidence, he seems to be appealing to similar ideas. Benoist takes Husserl to answer the “Twin Earth” question affirmatively (that the meaning of “this” differs), but, unlike Joumier and Beyer, Benoist seems to provide a more externalist reading (though whether the interpretation is ultimately externalist or not remains questionable). According to Benoist, the difference in meaning is due to the two distinct contexts (the surrounding-appearances that are seemingly identical) not having been “placed in the same way in the same world” (Benoist, 2002, p. 126). Benoist suggests that by means of the twin example Husserl ends up emphasizing “worldliness as constitutive of meaning,” which means that meaning is inseparable from “the world in which it appears” (Benoist, 2002, p. 126). Benoist does not cite further evidence in support of this reading of the Twin Earth passage, but I think there is such evidence in Husserl’s later texts.

In many texts from the 1920s and the 1930s, Husserl discusses the meaning of the world. In “Phänomenologie und Anthropologie” [1931], Husserl writes about “the real world” that “holds true for us, shows itself to us, the only world that has meaning for us” and how phenomenology investigates “the sources for that world’s meaning and validity for us, the sources that comprise the true meaning of its being” (Husserl, 1989, p. 179/498). In Sect. 4.1, I cited Husserl’s description of phenomenology, in the 1930 “Nachwort” to Ideen I, as clarifying “the meaning of this world, the precise sense in which everyone accepts it, and with undeniable right, as really existing,” but Husserl continues to say, and this is important, that “[t]he world has this meaning whether we are aware of it or not” (Husserl, 1930, p. 563/22). Husserl repeats this idea in his Encyclopedia Britannica article on phenomenology [1928], writing “that what belongs in and for itself to the world, is how it is whether or not I, or whoever, become by chance aware of it or not” (Husserl, 1962b, p. 288/169), as well as in “Amsterdamer Vorträge” (1928) where he writes that “everything which belongs to the world is ‘in and for itself’ as it is, whether or not I, or whoever, may be accidentally aware of it or not” (Husserl, 1962c, p. 332/239). “For, in advance,” Husserl writes in Die Krisis der europäischen Wissenschaften und die transzendentale Phänomenologie [1936], “’world’ has the meaning ‘the universe of the ‘actually’ existing actualities’” (Husserl, 1976b, p. 148/146). Since the world has this meaning “in advance,” Husserl seems to appeal to the same idea of meaning that partly escapes the subject.

There are two important things in these passages. First, Husserl repeatedly affirms that the world has either the meaning or the property of being “in and for itself” irrespective of whether anyone is aware of it. Husserl thus brings the idea of the identity of meaning, which can be subjectively indistinguishable, to the world. Second, Husserl talks about the world’s independent existence as both something the world has and as its meaning. One might here see a distinction between discourse about the property of the world and the meaning of the subjects’ experiences of the world, but the nature of Husserl’s language implies the crucial idea in his philosophy that seems to challenge such a distinction: the correlation.

Husserl characterizes his phenomenology as “correlational research” (Husserl, 1973b, p. 121/88), which, in the broadest sense, means investigating the correlation between consciousness and objectivity. More specifically, phenomenology investigates the correlation between intentional objects and intentional acts. In brief, the idea of the correlation is that the mode in which the object is given, or the way it appears to consciousness, is conditioned by both the act that intends the object and the object that is given to the act. One of the starting points of Husserl’s phenomenology is to overcome the natural way of grasping objects as utterly independent from consciousness, via the epoché, and rather consider objects in their modes of givenness. Given the sense of the correlation, however, it is equally important for phenomenology to overcome the way of grasping acts as utterly independent from objects and thus consider acts in correlation with their objects.

One general way to grasp the correlation is to think of it as a relation of mutual dependence. For example, the correlation can prevail between an act and an object in which both depend on one another. However, the nature of this dependence is a more specific question. In his book on continental realism and anti-realism, Sebold distinguishes between different alternative ways of reading Husserl’s concept of the correlation, two of which are relevant in this context. On the one hand, correlation can be understood as metaphysical correlation. If consciousness and the world were in such metaphysical correlation, the existence of the world would depend on the existence of consciousness (Sebold, 2014, p. 185). This is indeed how Smith seems to understand Husserl. As said above, Smith claims that Husserl is “an absolute idealist” to whom nothing would exist if consciousness did not exist (Smith, 2003, p. 179). However, correlation can also be understood, on the other hand, as semantic correlation. Sebold explains the idea of semantic correlation in the following way: “The physical world itself is not dependent on consciousness; it remains an unproblematically independent existent. The meaning of the physical world, on the other hand, is relative to subjectivity; there is no meaning without consciousness.“ (Sebold, 2014, p. 199.) Semantic correlation indicates that the meaning of the world (as a real world), rather than its existence, depends on consciousness. Although the term “semantic” might give the incorrect impression that Husserl is only concerned with meaning in the linguistic sense, I think the concept of semantic correlation captures the key idea of Husserl’s correlation insofar as “meaning” is understood more broadly (see footnote 4). However, I think it is important to emphasize, in light of the preceding presentation, that the dependence relation of meaning goes both ways: not only does the meaning of the world depend on consciousness; it also depends on the world.

Husserl’s concept of constitution is important here because constitution is the process that enables the world to be meaningful, i.e., to manifest as it does (as a real world). Husserl also understands constitution in correlational terms (see Husserl, 1952, p. 179/189). Like Crowell explains, consciousness allows the world to manifest as it does, but consciousness could not do this if there was no world to be enabled to manifest as such. World and consciousness depend on one another. Therefore, “the meaning of the world” can be both something the world has and something our experience of the world carries. On the one hand, the meaning of the world as real is something that the world receives from consciousness. Yet, this is not to say that the world did not exist prior to or without consciousness; it is rather that the world would not have any meaning if there was no meaningful experience to disclose the world. In a word, the dependence between the world and experience is semantic, not metaphysical. On the other hand, the meaning of the world transcends the subject because the world has it regardless of the subject’s or other subjects’ awareness of it. The meaning of the world, and meaning in general, has a worldly character to it. The world is meaningful, and meaning is worldly. In “Amsterdamer Vorträge”, Husserl puts this explicitly:

We, individually and in cognitive community, are supposed to be the ones in whose conscious life-processes the real and every ideal world should gain meaning and acceptance according to all that they are (as pregiven to us, at hand, and as existing in and for themselves). We ourselves, however, as human creatures, supposedly belong only to the real world. In accordance with the worldliness of our meaning, we are again referred back to ourselves and the conscious life wherein this special meaning takes shape. (Husserl, 1962c, pp. 333–334/240)

As far as I know, this is the closest Husserl comes to placing the source of the identity of meaning. In using the expression “the worldliness of meaning” or “worldly meaning” (weltlicher Sinn), Husserl places that source partly in the world. Since Husserl talks about how we belong to “the real world” in the preceding sentence, it is evident that Husserl is referring to the real world rather than some veil of appearances. To Husserl, in fact, there is no such two-world distinction; the real world is the lived world of experience (see e.g., Husserl, 1989, p. 179/498). Therefore, by “worldly meaning,” Husserl highlights the roots of meaning in the real world. This is not to say, of course, that the source of the identity of meaning was only the world, which is not required by strong externalism either. The source is rather claimed to be in both the subject(s) and the real world.

The identity of meaning is thus no longer seen as stemming solely from the subject; its source has been placed partly in the real world, an idea that was already incubating in the 1911 manuscript. This means that meaning can depend on features of the real world. Yet the nature of this dependence remains to be articulated. Although Crowell sets his interpretation against standard externalism by replacing a causal relation between consciousness and the real world with a normative relation, it is worth emphasizing that externalism does not need to conceive of the relation as causal. In his foreword to an anthology of externalist papers Thought and Object [1982], Woodfield explains that externalist theories maintain that de re thought (i.e., thought about a particular object), “is about an actually existing object, and it is tied to that object constitutively, so that the thought could not exist without that very object’s existing,” but “[t]he kind of impossibility alluded to is logical or metaphysical rather than causal” (Woodfield, 1982, p. v). Woodfield continues by saying that “[t]he thought could not exist without the object because it is individuated in a way that makes its relatedness to that object essential to its nature” (Woodfield, 1982, p. v). While causation is one way to understand this relatedness, it is precisely the affirmation of such a dependence relation between the object in the real world and the mental state that is quintessential to externalism, regardless of whether the relation is causal or not.

When it comes to externalism, it seems that another property of the dependence relation is more important: whether it is symmetrical or asymmetrical. In arguing that externalism is philosophically significant because it entails realism, McGinn claims that externalism has this entailment because it grasps the dependence relation as asymmetrical. The relation could not be symmetrical, according to McGinn, because “what serves to individuate a mental state cannot itself be individuated by that mental state” (McGinn,1989, p. 10). “If an object’s being red consists in its looking red (crudely),” McGinn claims, “then it cannot also be true that its looking red consists in being brought about by red things” (McGinn, 1989, p. 11). The difference between my thought about water and my twin’s thought about water stems from the difference between H2O and XYZ, but the latter difference cannot stem from the former. If it did, then the difference between H2O and XYZ would be due to a difference between the mental states about them (even though H2O and XYZ are subjectively indistinguishable). This would lead to metaphysical idealism where the worldly features are constructed by the mind, which would be incompatible with strong externalism. I call this argument the asymmetry argument.

A Husserlian critique of the asymmetry argument would point out that the argument exemplifies the sort of thinking that Husserl’s phenomenology tries to overcome, namely the thinking of the object in complete isolation from consciousness. The objects are there, and they individuate the mental states about them, but the objects do not depend in any way on the mental states. Rather than thinking of the object in complete isolation from the act of consciousness to which it is given, Husserl’s phenomenology uncovers both in correlation with one another. Where the asymmetry argument relies on a distinction between an object looking red and an object being red, Husserl challenges this very dichotomy. “[I]f we attempt in this way to separate the actual object (in the case of outer perception, the perceived thing of nature) and the intentional object, inserting the latter in a real [reell] way as ‘immanent’ to the perception, immanent to the experience,” Husserl argues, “then we land into the difficulty that now two realities are supposed to stand opposite one another, while only one is at hand and possible” (Husserl, 1976a, p. 207/179). “The true object (…),” to Husserl, “does not lie in a mystical outside and in-itself – an outside completely transcendent to consciousness, in principle inaccessible to I-consciousness” (Husserl, 1962a, pp. 430–431).

Although Husserl abandons the appearance-reality dichotomy, replacing it with what we might call the lived world, there remains some distinction between what appears to the subject and what does not. This is not only in reference to the difference between what is immediately given to me and what is implied in the horizon but also the difference between what is generally given to me in the stream of my experience and what remains unnoticed. Having argued that the non-existence of the real world is “a material absurdity,” Husserl writes in Ideen I that “[o]bviously there are things and worlds of things that cannot be demonstrated in a definite way in any human experience, but there are simply factual reasons for this, grounded in the factual limits of this experience” (Husserl, 1976a, p. 103/88). Certain things go unnoticed due to my lack of knowledge, but nevertheless determine meaning, as Putnam’s example of beeches and elms shows. In fact, if one looks at the issue of internalism and externalism within transcendental phenomenology, i.e., from within experience, then this seems to become the crucial philosophical issue: whether features that are subjectively indistinguishable – factors at “the limits of experience” – can determine meaning. The internalism-externalism framework has sometimes been criticized from the phenomenological perspective for failing to recognize the wholeness of the transcendental dimension that cannot be captured with the internal-external divide (see e.g., Zahavi, 2004, 2008, 2017, pp. 118–120; Murchadha, 2003, 2008; Man-To, 2014), but, as I said in Sect. 2, the criticism only applies if the internal-external divide is understood in spatial terms. If one understands the internal-external divide as the divide between what is subjectively distinguishable and what is not, then the argument no longer holds. The way I see it, this is precisely the way in which the question of internalism and externalism emerges when looked at within transcendental phenomenology.

Given the passage from Husserl’s Ding und Raum concerning the identity of sense and Husserl’s ideas about the worldliness of meaning, the answer to the question seems clear: meaning can be determined by worldly features that are subjectively indistinguishable. Regardless of their indistinguishability, these features are nevertheless connected to phenomenological immanence. Yet, this does not designate that they were somehow detached from the real world. On the contrary, the fact that our experience is characterized by subjectively indistinguishable aspects implies that there is a real world “out there” that partly escapes our epistemic grasp. However, the real world “out there” is not a world “beyond our experience,” in the sense that it would be in principle inaccessible, but precisely the real world in which we always live and experience.

When the problem of internalism and externalism is approached from such a phenomenological angle, the dependence relation between meaning and the real world does not seem asymmetrical. It seems to be symmetrical in the sense of the correlation. From a Husserlian perspective, the object looking red, to borrow McGinn’s example, would not be a mental state that may be impacted by the object being red; both the object looking red and the object being red are rather part and parcel of the meaning of the object and the experience. The meaning unfolds in the correlation between the perceptual experience of the red object and the red object of the perception. By discussing the meaning of the world and the worldliness of meaning, Husserl proposes that meaning is not strictly tied to either the subject or the object; it rather lies in the correlation between them. The best way to conceptualize worldly meaning is not necessarily to describe it as “the meaning that can have a valid relation to the object,” though it contrasts with own-meaning or opinion, because worldly meaning is already in the world. It is in the world, its objects, and its states of affairs, but since there is no other world than the lived world (i.e., the world given in experience), the meaning’s being in the world also indicates its being in experience.

The upshot is that, from the phenomenological perspective, the philosophical significance of externalism does not require the kind of asymmetrical dependence that insulates the object completely from consciousness. What is needed is rather dependence on features that can escape the subjects' ability to distinguish them. If meaning can depend on such factors, then there is a real world, and thus some kind of realism is entailed by externalism. The kind of realism is nothing but the common-sensical idea that the existence of the real world is independent from the subject(s), which indicates that the real world might nevertheless depend on the subject(s) for some of its features. This is the only kind of realism that is required for externalism to maintain its philosophical significance. Even if the dependence relation was symmetrical to the extent that object and act mutually depended on one another, this would not violate the philosophical significance of externalism insofar as the mutual dependence would be semantic rather than metaphysical. Husserlian phenomenology stops at these limits of experience and does not affirm anything more about the nature of the dependence relation, whereas traditional externalism is seen to bear certain metaphysical presuppositions (see Tahko, 2015). However, this investigation has shown that the philosophical significance of externalism does not necessarily demand such presuppositions. In the externalist interpretation of Husserl that I have developed, realism is entailed, but it is not presupposed.

5 Conclusion

I have argued that Husserl can be interpreted as a strong externalist. Having distinguished between the other people argument and the real world argument, I defined the relevant positions: where internalism rejects both arguments, externalism can accept either just the other people argument (weak externalism) or both arguments (strong externalism). Only strong externalism can preserve what McGinn calls the philosophical significance of externalism because it entails realism about the world. After investigating the 1911 manuscript in which Husserl presents his own “Twin Earth” case, I claimed that while the manuscript can be interpreted in either internalist or externalist terms, the “Twin Earth” passage is better understood from an externalist perspective since it invokes a notion of meaning that can contain aspects which are subjectively indistinguishable. I then showed that the externalist interpretation of the passage connects to ideas concerning the identity of sense and worldly meaning in Husserl’s other texts. Where I saw Smith’s weak externalist interpretation of Husserl as losing the philosophical significance of externalism, I mostly agreed with Crowell’s externalist interpretation, though I recognized an ambiguity regarding the relation between meaning and reality in the interpretation. My interpretation tried to clarify this ambiguity first by exploring texts in which Husserl talks about “the meaning of the world” and “the worldliness of meaning” as well as, second, by assessing McGinn’s asymmetry argument for the philosophical significance of externalism from the Husserlian perspective. I argued that the dependence relation in Husserl’s account is correlational, but it can nevertheless entail a kind of realism and thus preserve the philosophical significance of externalism without asserting the asymmetry argument.

The conclusion that Husserl can be read as a strong externalist might be surprising. Perhaps the reason is that Husserl’s transcendental phenomenology seems to stem from a contrasting point of departure than strong externalism. Where externalism is often guided by the idea that we can learn more about consciousness by studying the real world, Husserl’s phenomenology seems to be guided by the opposite idea that by studying consciousness, we gain insight into the structure of the real world. This idea might give rise to associations with idealism and internalism, but in fact it is just an acknowledgement of consciousness as the natural starting point for doing philosophy. Rather than deeming phenomenology and externalism contradictory, this investigation helps make better sense of their seemingly contrasting points of departure. For to gain insight into the real world by investigating consciousness, reality needs to have already blended with consciousness in some capacity. The externalist interpretation of Husserl offers one way of grasping the nature of this relatedness.