Is the Hilina Pali ‘palaeomagnetic excursion’ becoming another example of the reinforcement syndrome? A comment inspired by Nawrocki et al. (2018) IAN SNOWBALL Using a combination of Optically Stimulated Lumines- cence (OSL) dating, mineral (rock) magnetism and pal- aeomagnetism, Nawrocki et al. (2018) suggest that the Hilina Pali excursion (reviewed by Singer 2014) might have been recorded in loess profiles from western Ukraine, and dated it to c. 18ka. Theirpalaeomagneticdataalsoreveal a discontinuous illuvial horizon in Holocene soil with reversed characteristic remanences (ChRMs), which they state ‘result most probably from a self-reversing mecha- nism’.Rightly,theyconcludethatmoreinformationabout how pedogenic processes influence the acquisition of nat- ural remanent magnetizations (NRMs) is needed, specif- ically because our understanding of how palaeosols become naturally magnetized is poor. Thus, this com- ment on Nawrocki et al. (2018) does not focus on a self- reversing mechanism that might have produced reversed polarity ChRM directions in the palaeosol. Instead, it focuses on their report of ‘the first occurrence of the Hilina Pali [excursion] in a loess sediment’. Nawrocki et al. (2018) sampled an approximately 11- m-thick loess-palaeosol section (Rivne ‘A ’) at 5-cm vertical increments, producing standard sized palaeo- magnetic cylinder samples (25922 mm). They found that a single sample (R117A) of loess located at a depth of 6.55 m with a steep negative ChRM inclination of À78° and an easterly declination (approximately 90°). Moti- vated by this result, the authors undertook a stratigraph- ically denser sampling of a thinner 1-m-thick profile parallel to Rivne ‘A ’ to test the result. This sampling (Rivne ‘B’) was done 1 m away from Rivne ‘A ’ but with a vertical sample increment of 1.25 cm, which required that sampleswere obtained collaterally, with a horizontal offset of 5 cm. Again, they found a single sample (R33B) with steep negative inclination (À66°) in the second sequence, although the declination was about 355°. Based on these two samples with steep negative inclina- tions, which are surrounded by neighbouring samples of exclusively normal polarity (and no intermediate direc- tions, even in the more densely sampled profile), Nawrocki et al. (2018) conclude that there must have been recording of the Hilina Pali excursion in isolated domains of loess due to discontinuous deposition. It must be pointed out, however, that the authors recog- nized a series of five correlatable features in inclination and declination above and below the two reversed samples, which implies that discontinuous deposition is coincident with the proposed excursion. The oldest OSL date obtained by Nawrocki et al. (2018) is c. 41 ka and, if this age is accurate, one could reasonably expect to see some signs of the established Mono Lake (32 ka) and Laschamp (41 ka) excursions (Singer 2014), yet the palaeomagnetic data set from Rivne ‘A ’ shows no sign of them. One has to question the accuracy of the OSL ages, but at the same time admit that low mass accumulation rates, discontinuous deposition and natu- ral smoothing of NRM acquisition could preclude the recording of these relatively short duration excursions. Singer’s (2014) review of palaeomagnetic data during the Quaternary period concluded that the combined evi- dence for a global geomagnetic excursion at 17 ka is weak and controversial, although there are credible volcanic data from the Tianchi Volcano in China (Singer et al. 2014) that were not available for the review. This comment emphasizes that the global significance of a geomagnetic excursion is based on multiple palaeomagnetic data sets, each of which is continuous, and accurately and precisely dated (ideally by two independent methods that provide similar age–depth relationships). There are no grounds to reject the palaeomagnetic data published by Nawrocki et al. (2018), but the inadequate data set (two samples with steep negative inclinations but significantly different declinations) cannot be considered as credible evidence of the recording of a geomagnetic excursion in western Ukraine at c. 18 ka. The Hilina Pali excursion (note that the name Hilina Pali/Tianchi excursion was later adopted by Singer et al. 2014) is at risk of becoming another ill-fated example of the reinforcement syndrome (Watkins 1971). Reinforce- ment and the search for short-lived events plagued the palaeomagnetic community in the 1970s when, for example, there was a high-profile debate (Thompson & Berglund 1976) about the recording of a Late Weich- selian geomagnetic reversal in sediment sequences that were deposited in southern Sweden at c. 12 ka (e.g. M€ orner & Lanser 1974; N€ oel 1977). Four subsequent decades of palaeomagnetic research, which have seen the DOI 10.1111/bor.12328 © 2018 Collegium Boreas. Published by John Wiley & Sons Ltd BOREAS COMMENTS Boreas, Vol. 0, pp. 1–2