What is it that makes a coffee table and a kitchen table both tables? What is it that makes a Golden Retriever and a German Shepherd both dogs? What is it that makes you and I both human? Table, dog, human: these are all concepts used to unite many particular things into singular categories. But what are those concepts based on? What is it that makes it proper to classify a German Shepherd and a Golden Retriever as dogs, while excluding a squid from that same category?
This is the "age-old" philosophical conundrum known as "the problem of universals" ("universal" being the term which refers to concepts such as table, dog, and human) -- but what you will see very shortly is that there is not really any problem at all - and there never should have been.
The Battle: Realism vs. Nominalism
Historically, this "problem" of why we classify things the ways that we do has caused a philosophical rift between two camps: that of realism and that of nominalism. Realism holds that the similarities between particular objects are part of objective reality; that there is some real 'dog-like' essence to both Golden Retrievers and German Shepherds. In contrast, nominalism holds that those 'similarities' are only nominal -- in name only; that our classification of Golden Retrievers and German Shepherds into the category of dog is based, not in real similarities between the two, but in our subjective and pragmatic decision to classify them as such.
When it is broken down like that, it seems pretty obvious which position is the most conducive to objectivity. When simply stated as it is above, realism is the only rational option -- and nominalism is basically synonymous with subjectivism. Unfortunately though, there have been some very annoying and distracting ideas attributed to realism which have given it a far less attractive rap -- and those distracting ideas have arisen because of a failure to distinguish between three very different issues.
Three Separate Issues
When dealing with this topic of universals, there are three separate issues which have historically not been treated as separate -- and thus have led to mass confusion.
Issue #1: Metaphysical or Not? The first (and most foundational) issue to be addressed in this debate is that of whether or not the universal (essence, or similar attribute) is metaphysical, or not. That is to say that we must determine whether 'dogness' or 'man-ness' is part of metaphysical (objective) reality -- or whether is it only part of our subjective understanding of that reality. Realism says that universals are metaphysical. Nominalsim says that they are not. This issue is the essence of realism.
Issue #2: Metaphysical Nature. If, as the realist claims, universals are metaphysical, then they must have some sort of metaphysical nature. Are these universals 'perfect forms up in heaven' (Plato's theory), 'attributes intrinsic to particular things' (Aristotle's theory), or 'ideas in the mind of God' (Augustine's theory)? Or, are they something else which has not been discovered yet!? The point here is to understand that the issue of their metaphysical nature is not the same as the issue of whether or not they are metaphysical, per se. If Plato's theory that there is a perfect table up in heaven is false, this doesn't mean that his idea that 'table-ness' is metaphysical in some way is also false.
Issue #3: Epistemological Discovery. These universals, or concepts, are things which we use everyday in our reasoning about all of reality. If universals are metaphysical, we must ask ourselves how we come to discover them. How do we discover the metaphysical similarities between German Shepherds and Golden Retrievers, in order to arrive at the concept of dog being applied to both? Is it because we are remembering that similarity from a past life (Plato's theory), or is it by some other form of passive intuition (Aristotle's - and many others' - theory), or is it be some other means altogether!? I'll answer this one below, but the point here - as with Issue #2, above - is that this is a separate issue from the issue of whether or not the universals are metaphysical. Plato and Aristotle and every other philosopher since then can be (and many are) completely wrong about issues #2 and #3, but right about issue #1.
The idea that universals are metaphysical (i.e. real) does not in any way demand that there is a perfect dog or table up in heaven (Plato); nor does it demand that the only way to discover those universals is by some mystical experience or passive intuition (Plato, Aristotle, and many others). In other words, realism, properly defined and understood, only refers to the first issue: whether or not universals are metaphysical. Any given realist may have varying theories about the second and third issues (which will be either true of false), but those will be varieties of realism -- not realism, per se.
Missing this extremely obvious point is the only reason for the rabid rejection of realism by nominalists -- and surprisingly by Objectivists and Christians, alike. I'll leave the Christian aversion to realism for another day. Right now, I want to discuss the Objectivist 'response' to realism -- if you can call it that.
Objectivism's 'Response' to Realism (or 'Intrinsicism')
Objectivism, as an explicitly held philosophical system, begins with issue #3 (from above), and views everything else in metaphysics and epistemology through that lense. This is likely because Ayn Rand absolutely dominated every other philosophical thinker on that issue; on the issue of how we discover similarities between particular things in order to form universals or concepts in our minds. This was done in her theory of 'concept formation' which gives a remarkable account and defense of how we form various concepts based on sense perception.
Rather than believing that man discovers universals by mystical intuition or remembrance of past life experiences, Ayn Rand held that man discovers universal attributes and forms concepts through a process of abstraction (this theory is detailed in her 'Introduction to Objectivist Epistemology'). Thus, her theory of Issue #3 from above; her theory of discovering universals and forming concepts is rationally superior to that of Plato, Aristotle, and many others, who believed that universals were discovered in some way apart from reason.
However, as noted above, that issue of discovery is a very separate issue from the issue of whether or not universals are metaphysical -- or real (part of reality). Rand, Peikoff, and Objectivist thinkers in general have all confused this third issue of universal discovery and concept formation to be the end all and be all of the topic of universals in general -- when in reality, it is a mere sub-topic of the actual issue: whether or not universals are metaphysical. So Rand rejected realism based on non-essential ideas attributed to it -- and then, likely because of wanting to sound "pro-reality", she renamed that which she was denouncing: changing 'realism' to 'intrinsicism'. Since traditional realists (or 'intrisicists' to her) have been wrong on that third issue, Objectivists presume that realism is wrong in general -- without ever pausing to consider the alternatives or the implications. Believing themselves to have 'solved the problem' between realism and nominalism, they have really just taken positions on both sides in the attempt to have some sort of third option -- but there is no third option.
Real or Not: There is No Alternative
Rand's theory of concept formation was great at describing how we discover similarities between things -- but notice that it is about discovering something (i.e similarity) in reality. Regardless of how we discover them (and I think Rand is right about how we do), are the similarities between a Golden Retriever and a German Shepherd real or not? Are those attributes of dogs which make them dogs and not caterpillars, real or not? If they are real, then they objectively are - apart from any subjective discovery of them. If they are not real, then there is no objective basis for our classification of them into such categories.
In OPAR (Objectivism: The Philosophy of Ayn Rand), on page 143, Peikoff criticizes nominalism because, in it "there is no metaphysical basis for classification". Well, does Objectivism hold that there is a metaphysical basis for classification or not? If yes, welcome to realism! If no, welcome to nominalism. But those are the only two options. Either attributes like 'redness', 'man-ness', 'dog-ness', 'table-ness', etc... are objectively part of metaphysical reality, or they are subjective figments of our imaginations. Either reality is what is is "in itself", apart from any subject, or reality is not objective at all. The 'Objectivist' cannot have it both ways here. Either universals are real (i.e. objective ), or they are not real (i.e. subjective ). There is no third option.
No amount of changing the subject (by focusing on issue #3 from above), and no amount of fear about what the answer to issue #2 might be (the Objectivists is naturally afraid of the idea of non-physical reality) will change the fact that objectivity demands metaphysical realism. And now the Objectivist will have to choose: Will he embrace metaphysical realism as the only metaphysical foundation for an objective worldview, or will he - for the sake of protecting sacred cows in his worldview - forsake objectivity altogether and dive headlong into nominalism. Consciously or sub-consciously, everyone will do either one or the other. Which will you do?