An objective correlative is a literary term describing a set of situations, facts, and a series of events, that permit a reader to rebuild or experience in their own mind an emotion the author has in mind. It might not be too imprecise to try to approach this concept, if it should appear to be tricky to grasp, by picturing one or more of the following contexts:
In a crowded pub early on a Friday night, an observer surrounded by their friends finds themselves looking across curiously at a stranger sat on a table on their own. They play on their phone a little while looking a little uncomfortable as they wait either for their friends or for food. A little while later the observer turns again to see them thank the waiter for a burger, perhaps a burger a number of their friends turns to admire and discuss, picking up the menu as they do so. For our observer the person is more interesting. They cannot imagine what they are thinking. Have they moved into the area? Have they come back from somewhere to an empty fridge? Have they recently broken up with somebody, perhaps somebody who used to do the bulk of the cooking so that their own skills have waned in the intervening months or years and now it feels sad to go into the kitchen and slave away trying to make a shadow of what they were used to? However they feel about it they now proceed to eat and, soon enough, our observer turns again to see them eating as messily as it is possible to eat a large burger. There is ketchup down their chin. They put down their burger, wipe it off with a napkin, rub at their fingers a little, and take a sip of beer or whatever is their poison, and, looking out of the window at a tram going by and people walking across a pedestrian crossing behind it, variously reacting to the change to a don't walk red man; for a moment all of their discomfort is gone, there is only beer, or wine, or water, or tea, and a burger which looks delicious. Our lone diner now looks back at our observer, who turns back to his office party and tunes back in to the conversation around them enough to choose between a discussion of the health effects of airborne toner, and the cost of car repair.
We see a couple, a man and a woman, at a hockey game. The puck moves quickly. It goes behind the goal and is taken up to the other side. It looks like the man is living it the whole time. It looks at the same time that the woman is asking questions, not quite sure how to react. They are surrounded by mixed company, most of whom are dressed for the occasion as this man's girlfriend is not, but, tense now in anticipation, he does not pay much mind to her, and when a goal is scored, he is jumping up and down, hugging his friends, male and female, before, slowly, very slowly, after the loud jingles and riffs and announcements, and after the game is set up to go once again, he turns back to her, notices her, and tells her something perfunctorily. She looks a little awkward.
Are these examples of an objective correlative? Yes and no. Eliot seems to believe that such a context can be built which will with some certainty convey an emotion to an engaged reader. This is far from certain. Eliot believed culture was built upon Christianity. Even if this were so - to a large degree it may be said to be true enough of the canonical texts he tended to read, react to, write, and discuss - many people would have a very different idea of the term. In the first example above, it may be possible to figure the emotion and the lived experience of the observed lone diner at the moment they are looking out of the window. It seems clear enough that they are lost in a pleasant dining experience. This does not mean that we know how things feel before or after that moment. From the context, however, we might guess at the feeling of the observer. We might make various hypotheses depending upon whether they might be male or female, or something in between, whether they might be straight or gay (out or in the closet for the people they are with), whether the observed is male, or female, or something in between, attractive or otherwise. The problem is here that it doesn't stop there. Even if we can figure out the chord, so to speak, of that moment looking out of the window, a chord makes a great deal more sense as a part of a chord progression.
In the second example, once again, we might guess with reasonable accuracy at the feelings of the two spectators at various moments. We might, depending on our own experience, tend to feel sorry for the woman, resentful of the man, or indulgent of one or the other. We might want to see more of the other female spectators, or know more of the other male supporters. We might want to know the kind of conversation these two have in other social situations, or feel that we cannot yet begin to judge without seeing, say, the man sitting down to a romantic comedy with his girlfriend. We might want to know if they are brother and sister.
Now these might simply be poor objective correlatives. Certainly they are written by somebody who has reason to be suspicious of the term with its behaviourist and empiricist assumptions. Modernism for some set out with the idea that consciousness could be as good as uploaded and streamed into the mind. This might be doubted and leads to an ideologically laden, uneven relationship between author and reader. More recently, typically led by the short story and, often enough, drama, the reader is now given a role in imagining themselves into a situation and empowered in their subjective choices. In the last handful of decades, and indeed, in a strain of the short story that goes back as far as Chekhov and has often been popularised by consistently successful journals such as The New Yorker, the data points the reader is given are minimised so as to minimise the determinism of the emotional "formula". Readers may then create stories and characters and emotional states by discussing these stories. For an excellent example of how such a collaboration might work, readers might turn to the New Yorker Fiction podcast or indeed take a look at the reading group resources at a good independent publisher like Comma Press.