Considering any problem, one cannot avoid links (or at least allusions) to what other people have already written. This is a normal way of introducing new ideas into an existing cultural context, without which mutual understanding would be impossible. People do not merely tell each other what they think; they also give their partners as many conceptual clues as possible to facilitate the assimilation of the other’s ideas in the personal picture of the world. For a productive dialog, we need to discover some initial commonality, a preliminary zone of consent that could be extended in our communication. This is what references can do.
Of course, to serve as communication bridges, references must point to something already known to the parties. On the other hand, referencing a current generality is useless for the purpose, since there is no link to the particular topic. It may be difficult to find an optimal balance between commonality and specificity, and one often needs to get better acquainted with the partner to start any meaningful discussion.
Nobody can know everything at all. No person has read every book, examined every artifact, or experienced every turn of history. Luckily, this is not required for productive thinking, and the rediscovery of something that has already been many times discovered by somebody else is a natural mechanism of human development. Virtually, every child has to discover the world anew. On the other hand, the humanity is inhomogeneous, and different social groups cover different cultural scopes; what is commonly known and self-understood in one group may be quite exotic in another. Referencing often plays the role of a positioning mechanism, indicating a reference group.
Browsing links can also be useful to get more control over one’s own reflection. For example, if I see that somebody acts my way, I may be driven to get closer acquainted with that person, and possibly borrow some interesting ideas. Conversely, if I notice a coincidence with someone I don’t like, it may serve as either an indication of a lack of consistency in my thoughts, or a motif for my reassessment of the other.
With all that taken into account, introducing references in the conversation can be quite natural, desirable and productive. However, to be of any real use, referencing must follow certain simple rules. Thus, if a name is just mentioned without explaining how it is related to the topic, the reference, most likely, will be in vain. Many scientific publications suffer of this fault, their authors mentioning lots of irrelevant figures just "for completeness". This habit is being implanted by the traditional style of the "academic" journals, where the assessment of submissions largely depends on the list of references. This style can penetrate live conversation as well, and one can often hear how the interlocutors compete in pronouncing names, leaving almost no time for meaningful conversation.
Yet another instance of the bad style is "abusive" referencing, just calling somebody names. References of that kind are thrown to the partner as abstract labels, without much concern about clarity and rationality. What use calling somebody Wittgensteinian? Please explain, what exactly in that person resembles you Mr. Wittgenstein, and in which of his hypostases. And as soon as you have explained it, the very need of mentioning Wittgenstein has been eliminated. The true meaning of any abusive reference is expressed in four words: "I don't like you." Abusive referencing is often used to overwhelm the partners, to suppress their opinions, to belittle them. In fact, such referencing can only demonstrate one’s utter incapability of coherent thought, becoming a kind of self-abasement.
One more special case: excessive quoting. In this style, references may be quite relevant, and their links to the theme of discussion clearly transparent, but their quantity exceeds the "critical mass", after which there is no practical need in any more examples. If an assertion has already been supported by enough reference material, further accumulation of the possible interconnections becomes annoyingly useless.
A well balanced style of referencing must be adequate, friendly and moderate, so that the partners could enjoy their conversation and help each other in developing their personal views.