Like so much that is eternal in philosophy, the question “Why is there something rather than nothing?” is a deification of something some jerk said said once.
This question, and the principles it motivates, has been seen as an antisceptical stone on which Occam’s razor has blunted. However, it should be noted that Leibniz himself saw it as a whetstone for Occam’s razor. As Leibniz says in Theodicy
I think the responses to the question can be divided into four categories.
There is the Descartes-Leibniz school, which sees the question as having a straightforwardly argument which founds ontology as an independent science.
The skeptical Voltaire-Hume-Kant school which sees the question as having only trivial answers insufficient for a new science.
The antiskeptical1 approach which includes Charles S Peirce, G E Moore and David K Lewis. This school moves away from demanding “justified true beliefs” toward demanding “justified reasonable beliefs”.
Heidegger, who is Heidegger.
I approach this question by breaking it up into “Why”, “is there”, “something”, “rather than” and “nothing”. Let’s go one by one.
First, “Why”. The English language contains about a dozen interrogative words: Who, whose, what, where, when, whence, whither, whether, why, how and a couple others. These words create an expected type of answer. For instance, “Where” is answered by current location and “Whence” is answered by location of origin. That is, “Where is Grant?” is answered by “Grant’s tomb.” and “From whence was Grant?” is answered by “Point Prospect, Ohio.”.
Of course, this is not the time or place to give a complete theory of how this typing could work or does work in practice. Instead, this is the time and place to narrow in on “why”. Unlike all other types that emerge from interrogative words in English, the type of answers to why questions - the type of aitias or explanations - don’t seem to have an independent meaning outside of answering why questions. There are independent theories of space and time to provide answers to where and when questions, there are independent theories of origin and process to provide answers to whence and how questions but not why questions.
Dennett calls approaches which allow why questions “Design Stances” and emphasizes - correctly - that taking this stance is methodological not ontic (i.e. there is no necessary assumption of a designer).
Leibniz takes the approach of objectifying this stance by thinking not in terms of design but in designability. He rejects as mere possibilia objects which could not be designed, such as a sealed glass cylinder of helium at thermal equilibrium with an inch of vacuum on top. Therefore in Leibniz’s hands, design is indirectly ontic.
Kant went on to explicitly restrict designability to things whose design can be conceived of - that is to say, things of ordinary experience. This approach of disconnecting designability from ontology is a variation of the argument that Leibniz himself was making, as noted above. The question whether Leibniz restricts out/allows in the right things and the answer is it depends on how you interpret the rest of the sentence.
The next sentence part is “is there”. This consists of the copula and an indexical, so here is a good place to compare Moore & Heidegger.
Moore and Heidegger are both reacting to the broader changes in society and the intellectual world: World War One, Darwin, Einstein and especially the rise of the new empirical psychology.
The philosophical issue raised by the new psychology is not that sensation and object are unrelated. Rather, the issue is that the theories of nerves and heat co-suppose one another. Mill’s analysis of matter etc. into possible sensations is a theory which is possibly true like any other, rather than a certain foundation.
First I will discuss how Moore returns to Plato. When one reads modern philosophy (roughly philosophy from Descartes to before Nietzsche), two lacks immediately jump out at the viewer.
One is the complete absence of the thoughts, opinions, words and actions of the ordinary person. The other is the disappearance of aporia, as already pointed put by Voltaire.
Moore’s return to Plato is a return to the Socratic method, the philosophical resurgence of ordinary language over theories of the objects of ordinary language.
The influence and interest in Moore’s articles is due to the optimism in method. Philosophers were, for instance, freed of the need to show the universe is or is not deterministic. They could explore varieties of determinism and indeterminism and see how the theories play in ordinary language.
More specifically, Moore’s argument “Here is one hand” etc. is designed to work because of the structure of indexicals. But it’s well known that the loose use of indexicals leads to ontological problems. “The tallest person in this room” forces one to admit a room and a person with special properties - ‘being this room’ and ‘being tallest’ - which are not features of the object in question. These impredicative properties are exactly what Kant objected to against Leibniz.
This objection only if one sees it as a foundationalism, but if one does not then one already has undone the possibility of an independent science of being, which is all Kant wants.
Okay, let’s move on to Heidegger’s return to Plato. For a man who had once declared the destruction of all metaphysics as his goal, on a technical level Introduction To Metaphysics is distinguished from The Enneads - “This veritably universal (Being) is in nothing; for nothing has proceeded from its existence.” - mainly by being not as well written.
But is this so odd? Plotinus’s theology is famously relentlessly negative. To Hen is always transcendental, never immanent. One often sees this as a contrast between philosopher’s gods and ordinary people’s gods. I think Plato would love this, given his obsession with engaging with ordinary opinions. Still, one wonders how Plotinus could hope for Socrates heaven where he never runs of of people to annoy.
What Heidegger adds is, like Voltaire, the conviction of Original Sin2: “In the sweat of thy face shalt thou eat bread, till thou return unto the ground”.
Whether one is a Heideggerian is partly determined by whether one interprets original sin onticly or methodologically. Interpreted methodologically, this clearly points towards the Questions Concerning Technology etc., but this leaves behind the original question.
So, the next sentence part is “something”. As Wilfrid Sellars pointed out, the English language has a wonderful ability to distinguish between ‘something’ and ‘some thing’. The latter is always quantified, the former is more complicated.
This is illustrated by the following sophistry: “3 is a number. Thus something is a number. Therefore, some thing is a number. That is to say, numbers are things.”. There may or may not be abstract objects, but this resembles no argument for them.
The reason why not is that the Therefore step is not justified. The sentences “There is something” and “There is some thing” are not synonymous.
Unfortunately, Leibniz’s question was in a language which lacks such an operation. Therefore it can be interpreted as, for instance,
“Why is there some thing rather than not any thing?” which could be answered by giving a reason for any object.
“Why is there x rather than not any thing?” which is answered by a function whose range is ends and domain is things.
“Why is there x rather than not x?” which is also answered by a function, but one which must sensitive to the contrasting part.
Voltaire and Leibniz both clearly consider the third question the correct one. Leibniz shows this by continually bringing up new instances - “Why is there evil rather than only good?”, “Why is there contingent truths rather than only necessary truths?”, etc.. Voltaire shows this by accusing the Optimists of equivocation between the third question and the more trivial second question.
Both Leibniz & Voltaire are both distinct from Descartes & Moore, who address the first question. As in the case of question words, it is helpful to distinguish the questions by the expected type of the answer. The first question can be answered by a sentence with no type parameters, a constant sentence. The third clearly requires a sentence of arrow type.
All this is to return to something already discussed: is the riddle of being ontological or methodological? The classical American Pragmatists, for instance, interpreted the question methodologically.
Since C I Lewis was a pragmatist, this is good a place as any to move on to “rather than” with its modal implications.
Leibniz, as seen above, introduces the principle of sufficient reason as a modal analog of the principle of contradiction. Leibniz affirms that truths considered nonmodally - such as “the ratio of 2 to 4 is the same as that of 4 to 8” - do not have aitias. In fact, Leibniz’s pragmatic interest in the question is, as already many times noted, to weed out mere possibilia from real possibilia. That is, to indicate the virtually true.
There is much to rehearse about modality, but this post has already passed the length limit twice. Suffice to say the interest in the question is related to one’s interest in modality.
I will therefore pass then to the final sentence part “nothing”. Perhaps nothing should be said, but I will try to be brief anyway.
Here two issues come into play: “Does “nothing” play a meaningful role in the question at all?” and “If nothing plays a role in the question, then is it “no thing” or “nothing”?”.
We have already seen the contrast between e.g. Moore’s answer to “Why is there anything at all?” and Leibniz’s “Why is there x rather than only y?”, so I will say no more on that here.
All I will say for now is that nothing as a singular object not only in philosophical theories but also in scientific language, through mereology.
Mereology is the study of parts. For instance, in topology, there are two ways that a surface can have a hole: it can have a boundary (a la Lewis) or it can have a positive genus. The hole with a boundary is a proper part of the surface but the genus is not a proper part. This is how, for instance, spaces described in general relativity can be nontrivial without boundaries.
Mereological theories can have a top (an object which has each object as a part) and a bottom (an object which is a part of everything) which can be meaningfully be interpreted as everything and nothing respectively, despite being objects rather than quantifiers. For example, Weyl’s concept of a Riemann Surface starts by enumerating the parts (function elements, chains of function elements, analytic forms, neighborhoods, etc.) and looking at the top element, the Riemann Surface itself. This top element is not a dispensable part of the theory, since the whole power of Riemann Surfaces comes from how the top elements generated by different charts are distinguished.
If The Nothing is an allowable object, may we further say “The Nothing Nots”? This has a superficial resemblance with the Russell-Quine theory of descriptions. That is, the theory that “A cat” means “A something which cats”.
But Heidegger’s theory is more than that. His claim is that verb “negating” is satisfied by an object (or activity) which can do the work of Plotinus’s aperion dyad. That could be true or false.
Well, there is no room left for more, so the last thing I will say is “no thing”:
I have always read Hume as an antiskeptic and the end of the linked chapter as droll, but I am bowing to tradition here as to not stack different types of confusion.
I do not mean to imply by this biblical quote to imply original sin is a biblical doctrine, which it is not. Rather, I am following Voltaire’s usage.