For demonstration purposes I was wondering about some very easy to grasp LC abstractions and I came to the idea of a function that simply eats its argument, and nothing more.
If you apply λx. (Yes no lambda term after the point) to an argument the abstraction reduces to nothing.
Although this isn't very useful for computation maybe, do you think the idea itself is valid or not?
Ok thanks everyone. Although I still like to tinker with the idea in my head, I understand it's simply not an option to use this throwaway function when dealing with LC.
Every application in LC returns an expression. There are no birds in the forest of Smullyan that can not sing.