Establishing dependencies during language comprehension requires access to previously encoded information. In the current study, we investigate the speed and accuracy of processing dependencies containing subject-verb agreement. Three speed-accuracy tradeoff experiments investigate the accessibility of number information across different amounts of interpolated material. We test the idea that marked features (such as the plural) are explicitly encoded and hence more available in memory when forming dependencies based on number (as in agreement) and more liable to engender interference when carried by inappropriate (non-agreement controlling) nouns, as suggested by prior studies of agreement attraction in comprehension (Tanner, Nicol, & Brehm, 2014; Wagers, Lau, & Phillips, 2009). The results suggest that features may be quickly shunted from focal attention even when fairly minimal amounts of material intervene been a subject and its verb. Furthermore, patterns of processing dynamics observed here pose difficulties for major accounts of agreement attraction advanced in the literature and we argue that they receive a more natural interpretation in terms of independently motivated constructs from basic memory research. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]