Forecasting earthquakes implies that there are time-varying processes, which depend on the changing conditions deep in the Earth's crust prior to major seismic activity. These processes may be linearly or non-linearly correlated. In seismology, the research has traditionally focused on mechanical variables, including precursory ground deformation (revealing the build-up of stress deep below) and on prior seismic events (past earthquakes may be related to or even trigger future earthquakes). Since the results have been less than convincing, there is a general consensus in the seismology community that earthquake forecasting on time scales comparable to meteorological forecasts is still quite far in the future, if ever attainable. The starting point of the present review is to acknowledge that there are innumerable reports of other types of precursory phenomena observable on the ground or in near-Earth space ranging from the emission of electromagnetic waves from ultralow frequency (ULF) to near-infrared (NIR) and visible (VIS) light, electric field and magnetic field anomalies of various kinds (see below), all the way to widely reported but never fully understood unusual animal behavior. These precursory signals are intermittent and seem not to occur systematically before every major earthquake. As a result they are not widely accepted, because no one could fully explain their origins. In addition, the diversity of these signals makes them look unrelatable, hampering any progress. In the first part, we review evidence for a solid-state mechanism based on decades of research bridging semi-conductor physics, solid state chemistry and rock physics, that is capable of providing explanations for the diversity of reported pre-earthquake phenomena. In fact, it appears that all pre-earthquake phenomena might be traceable to a single fundamental process on the atomic scale: the rupture of peroxy bonds via activation of electronic charges, electrons and positive holes, in rocks subjected to tectonic stresses prior to seismic activity. The positive holes are defect electrons in the O2- sublattice. They are unusual inasmuch as they are able to flow out of the stressed rock volume, into and through the surrounding unstressed or less stressed rocks. They form electric currents that travel fast and far, causing along the way a wide range of physical and chemical follow-on processes: electrical ground potentials, stimulated infrared emission, massive air ionization, radon emanation, increased levels of ozone, toxic levels of carbon monoxide (CO) and more. In the second part, we critically examine satellite and ground station data, recorded before a selection of past large earthquakes. Some of the phenomena can be directly related to the peroxy defect theory, namely, radon gas emanations, corona discharges, thermal infrared emissions, air ionization, ion and electron content in the ionosphere, and electro-magnetic anomalies. Of course there is a need for further systematic investigations, continuing statistical examination of the relevance and confidence levels of the observable precursors. Only then will the scientific community be able to assess and eventually improve the performance of earthquake forecasts.