INT J LANG COMMUN DISORD, XXXX 2018, VOL. 00, NO. 0, 1–15 Research Report Textual cohesion in oral narrative and procedural discourse: the effects of ageing and cognitive skills Sue Sherratt† and Karen Bryan‡ †Communication Research Australia, Rankin Park, NSW, Australia ‡Sheffield Hallam University, City Campus, Sheffield, UK (Received December 2017; accepted October 2018) Abstract Background: Knowledge of the discourse performance of non-brain-damaged individuals is critical not only for its differentiation from disordered expression but also for more accurate models of ageing and communication. The effect of ageing and cognitive skills on the cohesive adequacy of discourse has, until now, presented a confusing and ambiguous picture. Aims: To examine comprehensively the effects of both age and cognitive skills on the discourse cohesion of 32 non-brain-damaged males divided into four age groups. Methods & Procedures: A large body of narrative and procedural samples (394 samples) was elicited from the participants. Their cognitive skills were determined using three tests, whilst their discourse cohesion was analyzed and correlated with the cognitive test results. Outcomes & Results: This extensive investigation of ageing effects on discourse cohesion and their relationship to cognitive behaviour did not provide neat generalizable results. It showed that ageing significantly increases the number of cohesive errors and reduces the quantity of referential ties in picture-sequence narratives. The changes with age were limited to two aspects of cohesion and not linear across age groups. The participants’ cognitive skills declined with age. Correlations between some cognitive tests and certain cohesive changes suggest co-occurring deficits rather than a causal explanation of cohesive decline with age. Conclusions & Implications: With ageing there are increased cohesive errors and decreased referential ties, co- occurring with declining cognitive skills. This study yields important guidance for future research, suggesting that picture-sequence narrative is the most effective tool for clinical evaluation of discourse, but also that findings from one discourse sample may be misleading. Keywords: Ageing, cohesion, discourse, attention, narratives, procedures, referential ties. What this paper adds What is already known on the subject Research into the effect of ageing on discourse has been limited and the effect of ageing and cognitive skills on the use of cohesion has presented an incomplete and ambiguous picture. Only two aspects of cohesive use have been widely reported as impaired amongst ageing individuals, namely disordered referencing and increased cohesive errors. Although it is suggested that age-related cognitive decline may affect discourse production, particularly the use of cohesive ties, research has been sparse. What this paper adds to existing knowledge The findings of this study, based on the analysis of a unique and extensive body of both narrative and procedural discourse data produced by healthy older males, have clarified some issues. With ageing, there are increased cohesive errors and decreased referential ties, particularly in picture-sequence discourse; these findings co-occur with, but Address correspondence to: Sue Sherratt, Communication Research Australia, Rankin Park, NSW 2287, Australia; e-mail: Communica- tion.Research.Oz@gmail.com International Journal of Language & Communication Disorders ISSN 1368-2822 print/ISSN 1460-6984 online C  2018 Royal College of Speech and Language Therapists DOI: 10.1111/1460-6984.12434