The central regions of nearby elliptical galaxies are dominated by baryons (stars) and provide interesting laboratories for studying the radial acceleration relation (RAR). We carry out exploratory analyses and discuss the possibility of constraining the RAR in the super-critical acceleration range $(10^{-9.5},\hspace{1ex}10^{-8})$~${\rm m}~{\rm s}^{-2}$ using a sample of nearly round pure-bulge (spheroidal, dispersion-dominated) galaxies including 24 ATLAS$^{\rm 3D}$ galaxies and 4201 SDSS galaxies covering a wide range of masses, sizes and luminosity density profiles. We consider a range of current possibilities for the stellar mass-to-light ratio ($M_\star/L$), its gradient and dark or phantom matter (DM/PM) halo profiles. We obtain the probability density functions (PDFs) of the parameters of the considered models via Bayesian inference based on spherical Jeans Monte Carlo modeling of the observed velocity dispersions. We then constrain the DM/PM-to-baryon acceleration ratio $a_{\rm X}/a_{\rm B}$ from the PDFs. Unless we ignore observed radial gradients in $M_\star/L$, or assume unreasonably strong gradients, marginalization over nuisance factors suggests $a_{\rm X}/a_{\rm B} = 10^{p} (a_{\rm B}/a_{+1})^q$ with $p = -1.00 \pm 0.03$ (stat) $^{+0.11}_{-0.06}$ (sys) and $q=-1.02 \pm 0.09$ (stat) $^{+0.16}_{-0.00}$ (sys) around a super-critical acceleration $a_{+1}\equiv 1.2\times 10^{-9}~{\rm m}~{\rm s}^{-2}$. In the context of the $\Lambda$CDM paradigm, this RAR suggests that the NFW DM halo profile is a reasonable description of galactic halos even after the processes of galaxy formation and evolution. In the context of the MOND paradigm, this RAR favors the Simple interpolating function but is inconsistent with the vast majority of other theoretical proposals and fitting functions motivated mainly from sub-critical acceleration data.

Modified Newtonian dynamics (MOND) provides a paradigm alternative to dark matter that has been successful in fitting and predicting the rich phenomenology of rotating disc galaxies. There have also been attempts to test MOND in dispersion-supported early-type galaxies, but it remains unclear whether MOND can fit the various empirical properties of early-type galaxies. As a way of rigorously testing MOND in elliptical galaxies we calculate the MOND-predicted velocity dispersion profiles (VDPs) in the inner regions of $\sim 2000$ nearly round SDSS elliptical galaxies under a variety of assumptions on VD anisotropy, and then compare the predicted distribution of VDP slopes with the observed distribution in 11 ATLAS3d galaxies selected with essentially the same criteria. We find that the MOND model parameterised with an interpolating function that works well for rotating galaxies can also reproduce the observed distribution of VDP slopes based only on the observed stellar mass distribution without DM or any other galaxy-to-galaxy varying factor. This is remarkable in view that Newtonian dynamics with DM requires a specific amount and/or profile of DM for each galaxy in order to reproduce the observed distribution of VDP slopes. When we analyse non-round galaxy samples using the MOND-based spherical Jeans equation, we do not find any systematic difference in the mean property of the VDP slope distribution compared with the nearly round sample. However, in line with previous studies of MOND through individual analyses of elliptical galaxies, varying MOND interpolating function or VD anisotropy can lead to systematic change in the VDP slope distribution, indicating that a statistical analysis of VDPs can be used to constrain specific MOND models with an accurate measurement of VDP slopes or a prior constraint on VD anisotropy.