We propose a method for measuring parity violation in neutral atoms. It is an adaptation of a seminal work by Fortson [Phys. Rev. Lett. {bf 70}, 2383 (1993)], proposing a scheme for a single trapped ion. In our version, a large sample of neutral atoms should be localised in an optical lattice overlapping a grid of detection sites, all tailored as the single site in Fortsons work. The methodology is of general applicability, but as an example we estimate the achievable signal in an experiment probing a nuclear spin independent parity violation on the line $6mathrm{s},^2mathrm{S}_{1/2}$--$5mathrm{d},^2mathrm{D}_{3/2}$ in $^{133}$Cs. The projected result is based on realistic parameters and textit{ab initio} calculations of transition amplitudes, using the relativistic coupled-cluster method. The final result is a predicted spectroscopic signature, evidencing parity violation, of the order of 1 Hz, for a sample of $10^8$ atoms. We show that a total interrogation time of 30000 s should suffice for achieving a precision of the order of 0.1% --- surpassing previous determinations of the weak charge in Cs by at least a factor of five.