When acceleration is constant, then distance is a quadratic function of time. So, the solutions can be complex numbers, right?
I can intuitively think of negative solutions like this:
If my scooter is at $x=0$ at $t=0$, it's the initial velocity is $-1ms^{-1}$ and it has a constant acceleration of $2ms^{-2}$, then the my displacement as a function of time is $x=t^2-t$. For, $x=2$, $t=2,-1$. I think of the negative solution, i.e. $-1s$ like this: If, when I started my scooter, time suddenly started to run backwards, then I'd end up at $2m$ after $1s$. So, negative solutions just answer the question, 'what would have happened $1s$ before?'.
But if this is the situation: My scooter is at at rest at $x=0$ at $t=0$, and I start it with a constant acceleration $2ms^{-2}$, then at what time is my displacement $-1m$? Clearly, I can never end up at $-1m$ at any instant, even if time starts running backwards, because $x=t^2$ only takes positive values even for negative $t$. The equation gives the value $t=i$. Can we think of this imaginary time like this?: If when I start my scooter, time suddenly started to run in a perpendicular direction, then I'll end up at $-1m$ after $1s$.