Solipsism is a huge problem for anyone interested in promoting introspection as a way to understand the mind (which includes me). You can only introspect on your own mind, not on anybody else’s. So technically, all you really know for sure is your own mind. The existence of any other minds is purely hypothetical.
The same would go for the existence of the entire world. If you accept introspectively known sense impressions as valid information, you realize that you have no other information. All your sensory data are known to you and only you, by mental impressions. A touch on the arm is known as the mental feeling of a touch on the arm. The arm itself knows nothing. All you can know for sure is the mental impressions you have of the world. You can’t know if anything else is really “out there.”
In the most extreme form, a solipsist asserts, “I am the only self that exists. All the rest of the world is, at best, a hypothesis, or possibly just a figment of my imagination.”
There is no way to refute solipsism. Any counter-argument against it would just be another figment of my imagination. If it is false, I could never know it, because my own mind is the only thing known to me. Solipsism is an extreme form of idealism, which says that only mental events can be known to exist (or, only mental events do exist).
Once you take introspective findings as valid knowledge, you are confronted with the question, How is introspective knowledge different from other, empirical knowledge, such as scientific knowledge? The difference is that introspective knowledge of one’s own mind is certain, whereas scientific knowledge is hypothetical, merely a set of agreed-upon propositions. Scientific knowledge cannot be certain because it is not acquired through introspection, which gives the only direct, certain knowledge. (Image: feelwelcome.com.uk)
Consequently, in scientific psychology, introspection is not allowed. No introspective observations can be accepted into discussion of how the mind works because introspection is private, and if you accept private data as valid, it takes precedence over hypothetical, consensus-based scientific data, and no further scientific agreement or progress can be expected or achieved. In other words, introspection implicitly carries the threat of idealism, and then solipsism, which is ultimately nihilistic. If my own mind is the only mind that can be known directly for sure, how is a scientific psychology possible? It isn’t. The threat of solipsism therefore is serious. It would destroy everything else. That’s why it is simply outlawed, and so is introspection. And that’s why there is no generally accepted methodology like “scientific introspection.” (Despite that, I have published a book by that title, explaining how it would be possible).
The threat of solipsism is false; not a real threat at all. It is based on a misunderstanding of the human mind, which does not, and cannot exist in isolation from other human minds. One's own self and mind are learned (acquired) from socialization and cannot ever be separated from that context. The image of Rodin’s solitary thinker is profoundly misleading. We are not monads, and never have been.
The philosophical problem of solipsism is posed by abstracting one’s own mind from that of others, but this abstraction presupposes that the world is already given as a shared world. Hence solipsism presupposes its own refutation. It is a confusion, not a valid proposition.
True solipsism would require that I do not experience myself as a single self in distinction from other selves, but that I experience myself as the only self that exists. But that is impossible, for self is only defined by other. So again, solipsism is impossible in principle.
What about a person, say, an infant, who has virtually no self-awareness. Could that person be a solipsist? Such a person does not have the resources to contemplate the possibility of solipsism. So the thesis of solipsism is impossible in principle in this case also.
Suppose a philosopher, using reason and analysis, abstracts the personal self away from its social origins and maintenance, and considers it as an absolute, transcendental ego, disconnected from all others. From that position of the abstracted transcendental ego, could solipsism be taken seriously?
Husserl, inventor of the transcendental ego, might seem to have believed that. But he also wrote that only his reflections on intersubjectivity make “full and proper sense” of the transcendental ego (Husserl cited by Zahavi, 1996). This is why Husserl claims that a phenomenological discussion of subjectivity in the end turns out to be a discussion not simply of the I, but of the we. Thus once again, even from the position of the transcendental ego, solipsism is not possible in principle.
What is possible: An object can be experienced in different mental attitudes. Hegel noted that a book can be experienced by the senses not as a book, but as merely an existent object with properties, not as a social, historical object with meaning. So it is possible to “pretend” or imagine that one’s own self is merely an existent object, divorced of its social meaning. But that is imagination. We can imagine flying pigs, too, but that doesn’t prove a thing. We can imagine an isolated, mondadic self, but to take that fantasy seriously is the delusion that constitutes solipsism. So that solves a problem you didn't even know you had. Don't thank me.
Zahavi, D. (1996). Husserl's Intersubjective Transformation of Transcendental Philosophy. Journal of the British Society for Phenomenology 27 (3), 228-245.