Science (from scientia, Latin for "knowledge") has come to mean a body of knowledge, or a method of study devoted to developing this body of knowledge, concerning the universe gained through methodological observation and experimentation. The scientific method consists of different principles and procedures that are useful in acquiring scientific knowledge. Exactly what constitutes science and scientific methods are subjects studied by the philosophy of science.
Science or rather scientific laws and theories rest upon certain assumptions. One principle in particular is induction. Induction is a presupposition that the sequence of events in the future will occur as they always have in the past or that things in the universe will always act in the same way under the same conditions. For example, by induction the theory of gravity implies that a rock, thrown into the air with less than escape velocity will always fall back down just as other objects thrown into the sky have done in the past. It is by the assumption of induction that a scientific law or theory derives its primary strength to predict future events and explain past ones. The philosopher David Hume first attacked this assumption in what is known as the problem of induction.
Implicit in science's devotion to acquiring knowledge about the universe is an assumption that there is a reality that exists independent of a mind (or minds) perceiving it. This view, realism, holds that the universe (atoms, animals, gravity, stars, wind, microbes, etc.) exists independent of our observation. Under this view, the (approximate) truth of scientific knowledge is taken at face value.
Some of the findings of science under this view can be quite extraordinary to a non-scientific mind in light of every day common observation. Atomic theory, for example, implies that a granite boulder which appears as heavy, hard, solid, grey, etc. is actually a combination of subatomic particles moving very rapidly in an area consisting mostly of empty space.
Philosophers sometimes distinguish between the actual reality of things within the universe, which may or may not be fully perceivable by humans, and our perception of things within the universe. Immanuel Kant coined the phrases phenomena (the universe as humans experience it) and noumena[?] (things-in-themselves).
Realism, however, isn't necessary to science. Instrumentalism[?], for example, posits that while entities, such as atoms, help explain and predict data from experiments, these entities do not necessarily exist. This approach is favored by some when it comes to committing to the ontological status of a scientific entity which may seem unobservable[?] in principle.
In contrast to Kant's views (and despite wide acceptance that human perception of phenomena isn't necessarily an accurate reflection of the universe as it really is), most scientists assert that it is possible to understand and accurately explain (at least somewhat if not fully) the universe using the scientific method to hone accurate scientific theories and laws.
Scientists point out that while some people criticise the basic ideas of science, it is science alone that has provided information on the mysteries of the atom, the cell, the solar system, and the observable universe. It is science alone that has provided knowledge to develop tens of thousands of technological advances in medicine, engineering, communications and beyond. No other system which claims to compete with science has ever actually succeeded in actually producing useful information about the physical world in which we live.
Until the Enlightenment, the word "science" (or its Latin cognate) meant any systematic or exact, recorded knowledge (and the word continues to be used in this sense sometimes). "Science" therefore had the same sort of very broad meaning that "philosophy" had at that time.
There was a distinction between, for example, "natural science" and "moral science," which latter included what we now call philosophy, and this mirrored a distinction between "natural philosophy" and "moral philosophy." More recently, "science" has come to be restricted to what used to be called "natural science" or "natural philosophy," and further distinctions have been drawn within it, such as physical science, biological science, and social science.
Fields of study are often distinguished in terms of hard sciences and soft science. Physics, chemistry, biology and geology are all forms of hard sciences. They rely solely on the scientific method. Studies of history and sociology are sometimes called "soft science".
Mathematics is widely believed to be a science, but it isn't. It is more closely related to Logic; it isn't a science because it makes no attempt to gain empirical knowledge. However, mathematics is the universal language of all sciences.
The term "science" is sometimes pressed into service for new and interdisciplinary fields that make use of scientific methods at least in part, and which in any case aspire to be systematic and careful explorations of their subjects, including computer science, library and information science, and environmental science. Mathematics and computer science reside under "Q" in the Library of Congress classification, along with all else we now call science.
One of the key differences between religion and science is that scientists are willing (and sometimes, enthusiastic) to change their beliefs when new facts and compelling logic are presented. This subject is discussed in the article The relationship between religion and science.
Organization and practice of science: International Council of Science (ICSU)