The brain is highly interconnected (Schuez and Braitenberg 2002). Each neuron receives input from up to 10000 other neurons. Most brain areas are connected to most other brain areas and all layers are connected to all other layers (Constantinople and Bruno 2013; Felleman and Van Essen 1991; Harris et al. 2019). Thus, as far as the time scales of fMRI concerns, it can be challenging to treat different parcels of the brain (brain areas, layers, columns) as independent entities. In fact, even before the onset of the fMRI signal (500 ms - 600 ms after stimulus onset), the first input into a specific layer of a given brain area can spread the signal across all cortical depth and brain areas. Before the fMRI signal reaches its peak amplitude (6s–10s after the stimulus onset) the neural signal might have travelled to multiple other brain areas back and forth multiple times already. While this fast interconnectivity theoretically challenges the interpretability of layer-dependent functional connectivity, it does not seem to be a limiting factor in practice. Several potential mechanisms have been proposed, why this might be the case. They are discussed in the following sections.
