Material Characterisation of 19–20th Century Manuscripts from Northern Thailand

2.1 Manuscripts Investigated

The four manuscripts (Figure 1) employed for this study are part of a larger group of written artefacts from Southeast Asia that were loaned to the Centre for the Study of Manuscript Cultures for scientific investigation. Three of them (referred to here as MS 3, MS 4, and MS 6) belong to Singkha Wannasai’s library in Lamphun, a province in Northern Thailand. It is noteworthy that Singkha Wannasai (1920–1980 ce) was a prominent scholar of Lan Na manuscripts and one of the pioneering lecturers on Lan Na language and literature at Chiang Mai University in the 1970s (Grabowsky 2019, xiii-xv). His private collection, which later became the main part of Singkha Wannasai’s library, consists of several Lan Na manuscripts acquired from different provinces in Northern Thailand. MS 7 belongs to Dr. Anant Laulertworakul (born 1968 ce), a former lecturer of Thai etymology, epigraphy, and Lan Na manuscripts at the Faculty of Arts within Chulalongkorn University in Bangkok, who acquired it from a laywoman in Lampang province (Northern Thailand). The selected manuscripts originate from Northern Thailand and are dated or datable ante quem to the 19th and 20th centuries, thus making for a convenient corpus for investigating the interactions between modern technologies and traditional manuscript culture in Lan Na.

Figure 1:

The four manuscripts (a) Thai MS 3; (b) Thai MS 4; (c) Thai MS 6; and (d) Thai MS 7.

Thai MS 3 (Figure 1a): The manuscript measures 27.5 × 23.5 cm and consists of 33 paper folios, each approximately 0.15 mm in thickness. Examination of the paper with the naked eye did not show any identifiable sieve structure. In conjunction with the date of the manuscript and unified characteristics within paper sheets, this suggests that the paper might be industrially produced. The folios were bound together by two staples in the left-side vertical edge, mimicking the stab-stitched form which is normally obtained by sewing the folios with a thread. In addition, what appears to be a split pin is found in the centre of the same edge, inserted from the back. The folios consist of a combination of bifolios and single folios: three quires can be clearly recognised (fols 1–7, possibly with one folio missing at the front; fols 15–22; fols 27–33, possibly with a folio missing at the end), and the holes of overcast stitching suggest that fols 11–14 were also tied together. The manuscript was mainly written in black ink in Dhamma script and organised in 18 lines of scriptio continua per page. Some of the pages have been ruled with pencil. On the top left corner of each page, there is pagination written in Dhamma script, appearing until page 34 (fol. 16v), although one folio is likely missing between fols 10 and 11, as the pagination is lacking numbers 21 and 22. This multiple-text manuscript, written by a single scribe, comprises three texts: a jātaka tale titled Indakumāra (fols 1r–15v), a general didactic text (fols 16r–22r), and the tale of Nang Pathawi (fols 23r–33v). The writing on the front and back covers is so faint that it is barely readable. Structuring paratext addressed to readers is sometimes found, written in black ink by the same hand as the main text, in central Thai script, although the language is Tai Yuan vernacular (Supplementary Material, MS 3 Annotation A–D). Furthermore, the scribe left several colophons using black and blue inks, and regular and coloured pencils. On fol. 15v, corresponding to the end of the first text, there are two colophons and two annotations dating to 1929 ce, written in different scripts, with different media, within the span of a few days (Supplementary Material, MS 3 Colophon A). On fol. 22r, corresponding to the end of the second text, there are two colophons, dating to 1934 ce, indicating that the second text had been written five years later than the first part. Between the two colophons there is an ambiguous mantra verse written in blue pencil that is possibly a magical spell and a short instruction of recitation. This combined annotation does not seem to have any connection with the main text and the colophons and might have been added later in a blank space (Supplementary Material, MS 3 Colophon B). On fol. 33v, the colophon corresponding to the end of the third text can be found. Unfortunately, this text is only partially legible and could not be dated.

Thai MS 4 (Figure 1b): This paper manuscript is in leporello form. It measures 19.5 × 9.5 cm and is made of six connected sheets of paper for a total of 120 panels or folded pages. ‘Folded page’ is the term normally used in Thai disciplines to describe and number the folded parts of the writing support and follows the order of reading of the text, from front cover to back cover on the recto and from back cover to front cover on the verso. This means that fol.1v is not on the back of fol.1r. To avoid confusion, we use the term ‘panel’ to refer to the folded portion of the paper sheet, including both recto and verso. The writing support, consisting of two types of woven paper made in a floating mould, presented with different thicknesses. No sieve structure was observed with a light box. Type 1 corresponds to the majority of the paper used in the manuscript and has a thickness of 0.27 mm–0.33 mm. The thinner paper type (type 2), 0.17 mm–0.21 mm thick, was used twice to insert or replace small parts of the text, for a total of three panels. The sheets of paper had been connected by pasting the extremities together. For the two inserted sheets, one of the extremities had been sewn with a green thread. In some cases, the folds had been damaged and then repaired with self-adhesive tape. The front and back leporello covers are made of a thick (approx. 40 mm), folded, double-layered card-like paper, coated with a red layer. This multiple-text manuscript comprises two texts: Kammaṭṭhāna (meditation), consisting of nine parts, and medical recipes. The main texts were written with different hues of black ink. The texts were written in Dhamma and Central Thai scripts and organised in five to eight lines of scriptio continua per page, following blind ruling, except for the portion of text on the inserted sheets, where the ruling was made with a greyish writing medium. There are two colophons referring to the two main texts, partially written with pencils, that mention different years of writing: 1855 ce for the part on the recto and 1851 ce for the verso side of the manuscript (Supplementary Material, MS 4 Colophons A–B), showing that the main texts were not copied on a single instance. Full-page illustrations sketched in black ink, and additional texts and notes (in Dhamma script) in black, blue inks, and pencil, are found throughout the manuscript, and were possibly added later in the remaining blank pages of the manuscript.

Thai MS 6 (Figure 1c): This palm-leaf manuscript measures 55.7 × 5.0 cm and consists of 15 folios, made of the leaves of the Talipot palm (Corypha umbraculifera), whose margins were gilded with red and gold pigments. A pair of wooden covers protects the manuscript: they measure 55.5 × 5.5 cm, and are painted in red, with decorative floral patterns in gold on the margins. The manuscript is bound by a red thread that passes through the pair of aligned holes of the covers and the leaves and wraps around the bundle. The writing on each leaf runs for five lines of scriptio continua from left to right with some space left on the second, third and fourth lines around the pair of holes. The text was inscribed with a stylus and the incisions filled with black paste. It contains a text titled Kammavācā or ‘formulas for higher ordinations’. The colophon was written in the Dhamma script on the first leaf (fol. 1r), stating that the Kammavācā manuscript was donated to an unnamed monastery by a laywoman in 1970 ce (Supplementary Maetrial, MS 6). On the top left corner of the same leaf, there is a Pali text that reads “idaṃ saṅghassa kathinadussaṃ” (‘This is [the text] of the Kathin for the monastic community’). The Kathin is a religious ceremony in which laypeople offer monks robes and other alms-gifts after the end of the Buddhist Lenten period, and is not directly related to the main text of this manuscript (Premchit and Doré 1992). This note suggests that this leaf may have been the last folio of another bundle of palm-leaves bearing the Kathin text. As this folio was mostly empty, it is possible that the scribe of Thai MS 6 recycled this leaf by incorporating it to the volume hosting the Kammavācā and wrote its colophon there.

Thai MS 7 (Figure 1d): This palm-leaf manuscript measures 48.0 × 4.5 cm and is composed of 18 folios, also made using leaves of the Talipot palm (Corypha umbraculifera), whose margins were gilded with red and gold colours. Unlike Thai MS 6, it does not have a cover. The manuscript was written in Dhamma script and organised in five lines of scriptio continua with space interrupting the second, third and fourth lines to provide space for the red thread that binds the manuscript. The text goes left to right and was inscribed with a stylus and the incisions filled with black paste. The manuscript contains a text titled Maleyya or ‘a story of Venerable Maleyya’ (known in Central Thai as Phra Malai). The first folio of this manuscript (1r), displaying in the middle of the recto only the word “jūjaka” in Dhamma script and in pencil, was likely added by mistake. In fact, the holes for the thread do not align with the rest of the bundle. In addition, “jūjaka” is the title of one of the 13 episodes of Vessantara Jātaka. Since both stories of Maleyya and Vessantara Jātaka are chanted in an annual religious ceremony, this folio was perhaps displaced from its original bundle and added to this one. According to the colophon, the manuscript was copied in 1903 ce (CS 1265) from an exemplar called Wing wòn dòn klang chiang mai (Supplementary Material, MS 7). The scribe was a novice named Abhijaya, who copied the manuscript in assistance to a senior monk named Nandā at Thong Man Tai monastery.

2.2 Analytical Methods

In order to analyse the black writing media, we followed the now well-established protocol developed by the Centre for the Study of Manuscript Cultures (CSMC, University of Hamburg) and Bundesanstalt für Materialforschung und -prüfung (BAM) for the characterization of black inks (Colini et al. 2021; Rabin et al. 2012). The protocol consists of three steps: an initial screening to identify the type of ink, performed using a handheld Dino-Lite USB microscope AD413T-I2V under white light (visible), ultraviolet (395 nm), and near infrared (940 nm) radiation, followed by Infrared Reflectography beyond 1,500 nm to confirm the presence of carbon and an in-depth analysis using spectroscopic methods, such as X-ray fluorescence (XRF) and Raman spectroscopy. The red and gold paint used to decorate the manuscripts was analysed with XRF, Raman spectroscopy and VIS-NIR spectrophotometry. Writing supports of palm-leaf and paper manuscripts were analysed using Diffused Reflectance Fourier Transform Infrared spectroscopy (DRIFT-FTIR). The papers of Thai MS 3 and 4 were also sampled and subjected to additional investigation with optical and digital microscopy. The analysis was performed on multiple points per folio, spanning around 5 to 6 folios throughout the manuscript.

3.1 Thai MS 3

Thai MS 3 was penned predominantly with black ink, with a few sections written in pencil and blue ink (e.g., fol. 15v; Figure 2a). The digital microscopy images of the black and blue ink showed a decrease in opacity under near-infrared radiation (Figure 2b). XRF maps and line-scan measurements of the black ink indicated the presence of iron and calcium in the ink, suggesting the use of iron-gall ink in the manuscript (Figure 2c). No characteristic elements were detected when examining the blue ink, suggesting a blue organic dye. Methyl violet dye was identified by Raman spectroscopy (Figure 2d) thanks to its characteristic peaks at 1634 cm⁻¹, 1429 cm⁻¹, 1386 cm⁻¹, 910 cm⁻¹, 809 cm⁻¹, 481 cm⁻¹, and 446 cm⁻¹ (Borba, Honorato, and Juan 2015; Zaffino et al. 2017).

Figure 2:

Thai MS 3 (a) fol. 15v, indicating the regions analysed; (b) Digital microscopy images under (i) visible (ii) ultraviolet and (iii) infrared illumination; (c) XRF spatial map indicating the presence of Fe in the black ink, including contributions from the ink on the other side of the folio; (d) Raman spectrum of the blue ink; (e) DRIFTS-FTIR spectrum of the paper.

The writing support was subjected to both microscopic investigation and characterisation using FTIR spectroscopy. The FTIR spectra (Figure 2e) showed peaks corresponding to cellulose (3434 cm⁻¹, 2900 cm⁻¹, 1430 cm⁻¹, 1374 cm⁻¹, and 1030 cm⁻¹), starch (1000 cm⁻¹ and 989 cm⁻¹), and carbonate (871 cm⁻¹ and 1410 cm⁻¹), suggesting that the paper had been treated with calcium carbonate and sized with starch (Calvini and Gorassini 2002).

Library records accompanying the book indicated the paper to have been made from khoi (Streblus asper) or another type of mulberry fibre. However, under optical microscopy, fibres with window-like pits and frequent swellings were found in abundance (Figure 3b and c), alongside with fibres with thin lumens that exhibited reticulate perforations (Figure 3d), which indicated that they were fibres of softwood pulp. Additionally, a few singular smooth fibres were found, likely artificial impurities from the industrial papermaking process. No fibres of khoi or mulberry could be found. Examination of the surface with a digital microscope showed a regular topographical pattern that pointed toward an industrial origin (Figure 3a).

Figure 3:

Optical microscopy result for Thai MS 3 (a) digital micrograph showing a regular pattern on the surface of MS 3; (b) and (c) window-like pits and frequent swellings; (d) wide fibres with thin lumens, exhibiting reticulate perforations.

3.2 Thai MS 4

Visual examination of the leporello manuscript Thai MS 4 suggested the presence of two black inks employed for the main texts (fols 13r–14r; Figure 4a). They bear different shades of colour, one paler than the other. The more recent additions on the manuscript, the notes in pencil and blue pen, as well as the illustrations, were not considered for the analysis. According to the results of the protocol, both inks are carbon-based, as they did not lose opacity in NIR radiation (Figure 4b and c) and clearly showed the peaks of carbon at 1365 cm⁻¹ and 1601 cm⁻¹ in their Raman spectra (Figure 4d). No additional trace elements were detected with XRF spectroscopy.

Figure 4:

Thai MS 4 (a) fols 13r–14r with the indication of the areas analysed; (b) and (c) Digital microscopy images of the inks with different hue under (i) visible (ii) ultraviolet and (iii) infrared illumination; (d) Raman spectra of the two corresponding inks: ink 1 (used in Type 1 paper) and ink 2 (used in Type 2 paper) ; (e) DRIFT-FTIR spectra of the two paper types: P4 (Type 1) and P9 (Type 2).

The FTIR spectra of the writing supports of MS 4 indicated the use of two different papers. Peaks corresponding to kaolin at 3691 cm⁻¹ and 3919 cm⁻¹ were present in the spectra of type 1 paper, the thicker paper that was predominantly used in the manuscript, while kaolin spectra were absent in type 2 paper, the thinner paper used for the inserted sheets (Figure 4e, with type 1 represented by spectrum P4 and type 2 by spectrum P9). Peaks corresponding to cellulose (3434 cm⁻¹, 2900 cm⁻¹, 1430 cm⁻¹, 1374 cm⁻¹, and 1030 cm⁻¹), starch (1000 cm⁻¹ and 989 cm⁻¹) and carbonate (871 cm⁻¹ and 1410 cm⁻¹) were common for both papers. Lignin peaks were not observed.

Under digital microscopy, the papers appeared uneven, presenting swirls and accumulations of fibres consistent with woven paper made in a floating mould (Figure 5a). They seemed to have undergone only a low degree of beating: on some folios, large fragments of plant material are still visible, and under microscopy, thick bundles of unseparated fibres occurred (Figure 5b). Under optical microscopic observation, the fibres in both MS 4 paper types were long and narrow, with transparent membranes, i.e., primary fibre walls, enveloping the fibres (Figure 5c). The fibres presented an abundance of phytoliths, occasional broadening, and frequent cross markings (Figure 5d). In this case, the library records stating that the paper was made of mulberry were confirmed by microscopic observation, as the dominant fibres present in both the writing supports were khoi (S. asper) fibres – although a few cotton fibres were also found. The presence of cotton was particularly noticeable in Type 2, adding another parameter for the differentiation of the two papers. The difference between the two types of paper in MS 4 appeared to be, beyond the presence of kaolin as a filler in Type 1 and the thicknesses, a matter of the proportion of mulberry to cotton.

Figure 5:

Optical microscopy results for Thai MS 4 (a) uneven paper surface with swirls and accumulations of fibres, typical of woven paper made in a floating mould; (b) thick bundle of unseparated fibres, indicating a low degree of beating (c) long and narrow fibres exhibiting loose transparent membranes; (d) phytoliths, fibre broadening, and frequent cross markings, typical of khoi fibres.

Examination of the red pigment on the leporello cover using XRF spectroscopy indicated the presence of a significant amount of mercury, pointing to the use of vermillion. This was confirmed by the characteristic peaks at 253 cm⁻¹, 287 cm⁻¹, and 343 cm⁻¹ in the Raman spectrum (Figure 6a). Examination of the cover using FTIR spectroscopy (Figure 6b) showed peaks at 3433 cm⁻¹, 2930 cm⁻¹, and 252 cm⁻¹ that can be attributed to O–H and C–H stretching vibrations respectively. The peaks at around 1700 cm⁻¹, 1560 cm⁻¹, 1460 cm⁻¹, 1245 cm⁻¹, and 1070 cm⁻¹ are indicative of lacquer, possibly urushiol, a type of oriental lacquer (Hao et al. 2017; Narita et al. 2023; Okahisa, Narita, and Yoshimura et al. 2019).

Figure 6:

Analysis of the red pigment present in the cover of Thai MS 4 (a) Raman spectrum; (b) DRIFTS-FTIR spectrum.

3.3 Thai MS 6

The ink of this palm-leaf manuscript (Figure 7a) was identified as carbon ink based on results from digital microscopy (Figure 7b) and Raman spectroscopy (Figure 7c).

Figure 7:

Thai MS 6 (a) fol. 2r with the indication of the area analysed; (b) Digital microscopy images under (i) visible (ii) ultraviolet and (iii) infrared illumination; (c) Raman spectra of the ink; (d) DRIFTS-FTIR spectrum of the palm leaf.

The FTIR spectrum of the palm leaf (Figure 7d) showed peaks at 2926 cm⁻¹ and 2853 cm⁻¹ corresponding to the stretching vibrations of the CH₂ group. The peak around 1738 cm⁻¹ corresponds to the C=O group and can be attributed to both pectin and hemicellulose. The peaks at 1600 cm⁻¹, 1514 cm⁻¹, 1175 cm⁻¹, and 833 cm⁻¹ could be attributed to lignin, while peaks at around 1650 cm⁻¹ and 940 cm⁻¹ are characteristic of pectin. Peaks corresponding to cellulose were observed at 3403 cm⁻¹, 2917 cm⁻¹, 1377 cm⁻¹, and 1026 cm⁻¹ (Boeriu et al. 2004; Chu et al. 2023; Geminiani et al. 2022).

This manuscript additionally shows decoration in red and gold colour on the margins of the palm leaves and on the wooden covers. The XRF investigation of the red pigment on the wooden covers (Figure 8a) indicated the presence of lead, chromium, and traces of molybdenum. Raman spectroscopy (Figure 8b) confirmed the use of molybdenum orange, featuring characteristic peaks at 824 cm⁻¹ and 1179 cm⁻¹ (Chua et al. 2016). The golden colour on the cover was determined to be brass, as clearly indicated by the increase in the amounts of Cu and Zn in the XRF spectra. In contrast, the golden decoration in the edges is characterised by the presence of gold (Au) (Figure 8c), while the red pigment showed no difference to unpainted areas in elemental composition when examined using XRF, indicating the use of an organic dye. Reflectance measurement of the red pigment suggests the use of lac dye or alizarin (Figure 8d). The difference in the red and golden pigments used for decorating the covers and the margins suggests that they were produced in separate manufacturing contexts.

Figure 8:

Thai MS 6 (a) XRF spectra of the support and the red pigments present in the decoration on the wooden cover and the edges; (b) Raman spectrum of the red pigment on the cover; (c) XRF spectra of the support and the gold pigments present in the decoration on the wooden cover and the edges; (d) Reflectance measurement (inset) and first derivative plot of reflectance of the red pigments and references (IRUG database), suggesting the use of either Lac dye or Alizarin for the red in the edge.

3.4 Thai MS 7

The ink of this palm-leaf manuscript (Figure 9a) showed no significant change in opacity under infrared radiation (Figure 9b). XRF analysis of the ink indicated the presence of trace amounts of iron (Figure 9c), although both Raman spectroscopy and IRR (Figure 10a) confirmed the ink to be carbon-based.

Figure 9:

Thai MS 7 (a) fol. 6v–7r with the indication of the areas analysed; (b) Digital microscopy images under (i) visible (ii) ultraviolet and (iii) infrared illumination; (c) XRF spatial map indicating the presence of Fe in the ink (relative intensity); (d) DRIFTS-FTIR spectra of the palm leaf.

Figure 10:

Thai MS 7 (a) IRR image of a section of the palm leaf; (b) XRF spectra of the support and the red pigment in the margins of the manuscript; (c) XRF spectra of the support and the gold pigment in the margins of the manuscript.

The FTIR spectrum of the palm leaf showed peaks at 2930 cm⁻¹ and 2854 cm⁻¹ corresponding to the stretching vibrations of the CH₂ group while the peaks at 1600 cm⁻¹, 1520 cm⁻¹, 1208 cm⁻¹, 1179 cm⁻¹, and 812 cm⁻¹ could be attributed to lignin. Peaks corresponding to cellulose were also observed at 3422 cm⁻¹, 2904 cm⁻¹, 1425 cm⁻¹, 1377 cm⁻¹, and 1047 cm⁻¹; the peak at around 1736 cm⁻¹ was associated to the presence of hemicellulose and pectin, and the peak at 1660 cm⁻¹ was attributed to pectin (Figure 9d).

While this palm-leaf manuscript does not have an accompanying wooden cover, its margins are decorated with red and gold. The XRF results of both the red and gold pigments showed the presence of barium, suggesting an undercoating composed of barium sulphate. The use of barium sulphate in the pigment was further confirmed by the peaks at 1187 cm⁻¹, 1086 cm⁻¹, and 985 cm⁻¹ in the FTIR spectra (Cortea et al. 2023). Lead was detected with XRF in the red pigment (Figure 10b). However further attempts to characterise the red pigment using Raman spectroscopy did not yield any results due to a high degree of fluorescence similar to Thai MS 6. gold (Au) was detected with XRF in the golden decoration on the margins (Figure 10c).