Should we then abandon scientific realism, and embrace constructive empiricism? Well, the realists are not going to be impressed. And here are two possible realist rejoinders. The first one would say that the notion of empirical adequacy doesn't really do when it comes to explaining the success of science. And van Fraassen‘s Darwinian reformulation of the no miracles argument doesn't really cut any ice. One thing is to explain why only successful theory survives, another thing is to explain what makes a theory successful. So scientific realist would claim that she has a story to tell about what makes the theory successful, namely that, the theories that survive are successful because they are true, because the entities that they postulated are real, and because what the theory says about those entities is true or approximately true. The theories that didn't survive, the theories that failed, failed because they were simply false, like the Ether theory or Phlogiston Theory or the Caloric Theory. So this is the first rejoinder against van Fraassen and the reformulation of the no miracles argument. The second rejoinder attacks van Fraassen’s distinction between the observable phenomena and the unobservable entities. This distinction has been at the center of a very voluminous literature which I won't have the time to cover today. But, philosophers have been concerned about the way in which this distinction is drawn or can be drawn. Because it seems to be too stringent as an empiricist criterion. Why should we not rely on our scientific instruments, be they an electron microscope, or a particle collider, to deliver us a reliable image about what there is? Why should we trust our naked eye more than our scientific instruments? So here I won't enter into the intricate details of the arguments pros and against the observable and unobservable distinction, but I do want to mention one prominent realism response to constructive empiricism and this is a response that has been formulated by two philosophers of science, Philip Kitcher and Peter Lipton. The response says that we are justified to believe in unobservable entities because the inferential path that leads us to such unobservable entities is one and the same inferential path that leads us to unobserved observables. So what is an unobserved observables? Well consider the many ways in which perfectly observable objects may go unobserved. None of us has ever seen a dinosaur, yet if you could travel back in time, we could in principle be able to see a dinosaur with our naked eye. So a dinosaur is a classical example of an unobserved observable. So people like Kitcher and Lipton have been arguing that we are justified to believe in electrons, atoms, DNA sequence, unobservable entities on the same grounds on which we are justified to believe in dinosaurs. So how do we know about dinosaurs? Fossil evidence is what paleontologists use to reconstruct the past history of our planet. From fossil evidence of this type, we can reconstruct some important information about the life of extinct marine species like this trilobites in the Paleozoic era: for example, we can come to know whether they swam, they moved on the sea-bed, whether they ate plankton, how many different genders they were, are geographically distributed and so on. But as fossils provide evidence for now extinct species, similarly one can argue, the Large Hadron Collider's outcomes can provide evidence for the elusive explosion, inferential path to the unobservable explosion is one and the same as the inferential path that leads us to the unobserved observable trilobites. Philosophers of science call this inferential path “inference to the best explanation”. The idea behind inference to the best explanation is that we infer the hypothesis which would, if true, provide the best explanation of the available evidence. Thus, we infer the existence of marine anthropods like trilobites because this is the best explanation for these fossils as we infer the exposed one as the best explanation for the sort of evidence coming out of the LHC. Namely, we choose from a pool of competing explanatory hypotheses the one that we regard the best, the one that if true would provide a deeper understanding of the available evidence. Inference of the best explanation is a powerful tool in the scientific realist toolkit. It shows that the scientific hypothesis that we choose and we are willing to believe tend to be those that would provide a deeper understanding, the best explanation of our evidence, and this is what we tend to do every day in life. For example, in medical diagnostics, when the doctor infers the hypothesis that best explain the patient's symptoms, as well as in astronomy and cosmology, where we infer that the, the universe is expanding, as the best explanation for the cosmological redshift in the spectrum of the stars. The realists would say, this is what science is all about. We don't rely necessarily on our eyes or technological instruments to believe in the unobservable entities but on the validity and robustness of our inferential practices informed by a wealth of experimental data in order to draw conclusions about how our universe is going to be like.