1. Introduction

So present in diachronic writing, but so little explored linguistically. This is the duality that we face when it comes to abbreviations and the studies dedicated to them as a linguistic object capable of mirroring information of the person behind the quill. It is in this panorama that our work fits, by proposing the use of abbreviations as a graphic clue to characterize the 18th century scribe sociolinguistically. Thus, we intend to discuss how the use of brachygraphic resources is intertwined with external and internal factors of the language and how they, consequently, can also act as a methodological tool for the identification of the literacy degree.

In recent years, Brazilian linguistic literature has given a special look to abbreviations, in a diachronic bias, with regard to their relationship with linguistic phenomena and, so far, relevant results have been obtained. Among these works, those of Barbosa (2013) and Carneiro, Oliveira and Lobo (2013) stand out, who, like our research, propose that such resources can be instruments for the social characterization of the writer; Duchowny, Coelho & Coelho (2014) who found that variations in the use of abbreviations may be related to the socioeconomic stratum of the scribe; and Coelho, Ramos & Duchowny (2015) who also identified different patterns in the use of abbreviations, both in relation to frequency and in relation to the type employed, depending on the social stratum of the scribe; and those of Chaves (2006), Cohen (2010), and Seixas (2018), who attest that the abbreviations act as graphic clues that reflect processes of linguistic change.

It is important to mention that the literature already discusses the physical-calligraphic characteristics as a heuristic instrument for the identification of levels of writing ability (Marquilhas 1996; Barbosa 1999; Mendes 2008; Santiago 2012; Barbosa & Lima 2019; Santiago & Carneiro, 2019). This research, however, differs from those mentioned, since the object of our investigation does not fall on a characteristic of the form, but on the use of resources capable of denoting the linguistic competence of the person who wields the quill.

Thus, considering the relevance of the proposed theme and the state of the art that involves it, the research reported here presents itself as a possibility to expand “the understanding of linguistic phenomena related to the use of abbreviations in diachronic texts, as well as to contribute to the panorama of linguistic studies of sociolinguistic orientation, to the development of methodologies for work with corpora and, also, to the knowledge of socio-historical-cultural aspects of scribes in Brazil and Portugal in the 18th century” (Seixas 2020: 24).

Starting from the assumption that there is a norm for the use of brachygraphic resources in the 18th century and that there are levels of complexity of knowledge required to employ such rules (Seixas 2020), the following hypotheses were formulated and tested:

  1. External factors, such as level of education, socioeconomic stratum and sex, and internal factors, such as the type of abbreviations and the complexity of the rule, influence the use of abbreviations.

  2. Abbreviations are, therefore, a methodological tool for the characterization of the writer’s literacy degree.

Our general objective was, therefore, to analyze the use of abbreviations in eighteenth-century missives of the Portuguese Language to test the main hypothesis that such resources may show social aspects of the scribe and his degree of literacy, thus allowing its sociolinguistic characterization, which would prevent documents whose sociolinguistic record of the scribe could not be reconstructed in archives from being excluded from corpora of variationist researches. The following specific objectives were added to this greater objective: (i) to verify whether the external variables (level of education, socioeconomic stratum and gender) and the internal variables (type of abbreviation and complexity of the rule) influence the use of abbreviations and, if so, which variable has the greatest influence; (ii) establish methodological criteria for the indirect sociolinguistic characterization of the scribe by means of the employed abbreviations.

To test our hypothesis, it is necessary to know the complexity of the norm for the use of brachygraphic resources in the period, a topic that we deal with in the next section. In the third section, the theoretical-methodological procedures adopted will be presented, followed by our analysis and our final considerations, a section in which we systematize the generalizations achieved and present future perspectives.

2. Abbreviations

2.1. The norm

As the abbreviation rule, proposed by Seixas (2020), served as a criterion for analyzing the scribe’s domain of the norm and, consequently, the level of literacy, we present how this systematization took place. The author made a survey of 41 grammars and style manuals from the 18th and 19th centuries in search of information about the abbreviations, in order to identify and systematize the norm of the period, a procedure that led to the identification of 48 rules, which were organized into three subcategories: (i) prescriptions about the structure; (ii) prescriptions on the typology and (iii) prescriptions on the grammatical functions of the abbreviations.1 Considering that the treatment of the abbreviations’ employment rules took place in a heterogeneous way in the works, Seixas (2020) gave them another classification, namely: categorical rules (do not vary) and non-categorical rules (subject to variation).

Next, considering only the categorical rules, the aforementioned author assigned them a sub-category to help identify the level of literacy of the scribe, which takes into account the complexity of the norm, that is, the specific knowledge of the rules necessary to unveil the abbreviations rules:

  1. General rules (level 1): absence or low level of inference; it is possible to use this type because of the possibility of copying or observation and without revealing any character of the abbreviation. Example: Use a dot as an abbreviation sign.

  2. Specific rules (level 2): those that require specific knowledge, whether phonetic, semantic or grammatical, for the inference of elements of the abbreviation; it is necessary to have knowledge of a specific grammatical rule or specialized knowledge about a certain area of knowledge, which cannot be inferred; therefore, some formal instruction or metalinguistic knowledge is required. Example: Use reduplicated initial to represent that the abbreviated word is in the plural form.

Chart 1 shows the classification of the rules according to their level of complexity.

Chart 1

Rules and level of complexity.

Use Rules Level
Amount Do not write letters that the word does not have. 1
Initial Use uppercase character. 1
Use lowercase character in the abbreviation of que. 1
Abbreviation sign Use period as an abbreviation sign. 1
Use period after the last character. 1
Use period before the superscript letters. 1
Use tilde above q as an abbreviation of que.
Ex.:
1
Simple initial Use uppercase followed by a period.
Ex.: D. = Dom
1
Use lowercase followed by a period.
Ex.: q. = que
1
Superscript letter Use superscript letter(s) at the end of the abbreviation.
Ex.: M.a = Maria
1
Treatment pronoun Use uppercase initial.
Ex.: VS.a = Vossa Senhoria
1
Abbreviate the treatment pronoun by simple initial.
Ex.: Vm = Vossa mercê
1
Numeric Do not use Arabic and Roman numerals in the same abbreviation. 1
Do not start the abbreviation with a letter and end with Arabic numerals. 1
Use Arabic numerals and superscript letters.
Ex.: 7br.o = Setembro
2
Use Roman numerals and superscript letters.
Ex.: Xbr.o =Dezembro
2
Special sign Use symbol(s) and/or letter(s) with specific meaning.
Ex.: % = porcentagem
2
Tironian note Use shorthand symbol.
Ex.: & = et; e
2
Tilde Use the tilde to indicate nasalization.
Ex.: = Bom
2
Use the tilde to indicate double m or n inside the words.
Ex.: Mariãna = Marianna.
2
Use the tilde to avoid the ambiguity of past tenses perfect past and present future. Ex.: louvaram (past)/louvarão (future) 2
Place the tilde above the letter before the one that was deleted. 2
Do not put the tilde in the initial character.
Ex.: Aña = Anna
2
Plural form/Superlative Use reduplicated initial.
Exs.: DD. = Doutores/SSmo = Santissimo
2
Genre Use the ending -o or -a in a superscript position.
Exs.: Snr.or = Senhor/Snr.a = Senhora
2
Surname Use the letter z to indicate that the abbreviated word is a surname ending in -s
Ex.: Roiz = Rodrigues
2

2.2. Social history in the 18th century

Social history gave us support for the interpretation of socio-historical-cultural phenomena since, in the case of this study, one of the main aspects of the social history of the investigated writers is the relationship between writing and social practices, as according to our hypotheses, external factors are intertwined with them and can, consequently, be identified, as well as serve as a basis for the characterization of the writers.

2.2.1. Writing acquisition and development practices

When it comes to education, it is essential to consider that practices related to writing and literacy and the way of measuring them change over time, which means that the skills of a literate and an educated person in the 18th century are diferent than today’s. This makes the definition of what an educated citizen in the 18th century would be more complex, a standard we defined, then, not only by analogy with works of contemporary researchers, but also from the social history of the writers. Therefore, with regard to writing acquisition and development practices, such as initial reading instruction, schooling and literacy, we consider the assumptions of Soares (1998; 2003) and Marcuschi (2001).

Among these practices, we highlight the treatment given to literacy, since one of our objectives is to identify the level of instruction of the scribe. Thus, according to Soares (1998), today there are criteria to assess the degrees of literacy, which results from an equivalence between level of education and the ability to make effective and competent use of reading and writing, that is, the relationship between number of school grades completed by individuals, or their level of formal education, and level of literacy. This criterion is based on the assumption that, having reached a certain level of education, the individual will not only have acquired the technology of writing, that is, he will have become literate, but will also have appropriated the basic skills necessary to use the social practices of reading and writing and therefore will have become educated. In addition, for the author, literacy must be analyzed “according to the specific social needs and conditions of a given historical moment and a certain stage of development” (Soares 2003: 80).

Marcuschi (2001), following the same perspective, states that “there is no single literacy. Literacy is not the equivalent of acquiring writing. There are ‘social literacies’ that arise and develop on the margins of the school, and therefore need not be depreciated” (Marcuschi 2001: 19). It is seen that the author conceives the literacy process as a set of practices, or literacies, whose amplitude can vary from the minimum appropriation of writing, such as the individual who is illiterate, but educated to the extent that he identifies the value of money, he knows how to do calculations, he uses means of transport, even if he does not write or read regularly, until a deep appropriation, as in the case of individual writers.

As it is imperative to consider the socio-historical context of the period under analysis in order to understand the literacy criterion most appropriate to this context, we are based on the vision of Tfouni (2001) and Mendes (2008), according to which it is possible to evaluate the degree of instruction/level of education of an individual in the 18th century by means of scales related to professional occupation and the socioeconomic stratum of the scribes. Thus, since equivalence relations can be established, it is necessary to deal with formal education degrees in the 18th century. Implanted by the Pombaline Reform in 1759, teaching in public institutions took place through the Royal Classes, which comprised (i) the level of Minor Studies, subdivided into Chairs of First Letters (classes in reading, writing and counting) and in Humanities Chairs (Latin, Greek, Rhetoric and Philosophy) and (ii) the level of Major Studies, restricted to the university (Silva 2004). The training in Major Studies (higher level of education) or in Minor Studies (lower level of education) allows us to identify two distinct macrolevels of literacy from the 18th century – which, therefore, serve as criteria for measuring literacy degrees.

2.2.2. Social practices

When it comes to social history, it is important to consider that endogenous factors could reveal much about the education and literacy of individuals in the 18th century. Thus, we deal with the socio-historical aspects of both sides of the sea to understand the historical context in which the letters were produced. According to Magalhães (2001), the Portuguese society of the Old Regime was strictly hierarchical and in it “differentiated levels of reading instruction correspond to gradually differentiated levels of social inclusion/participation” (Magalhães 2001: 441). The social position and the social occupation/function were, therefore, closely intertwined in that period, since the type of job performed determined the social position that the individual occupied in the society in which he lived.

We cannot fail to mention that, among the considerations above, there was a social group that was generally excluded: the female sex. Even with the formalization of teaching and the creation of educational institutions, women were unable to attend these spaces. Until the end of the 18th century, there were no schools for girls in Portugal, leaving the convents, retreats, or private initiatives, when in wealthy families, the task of instructing women. This group did not have literate knowledge as a generic characteristic, not even in the elite, due to their social role and restricted access to educational practices (Algranti 1993; Adão 1997; Magalhães 2001).

As our corpus is composed of documents written by both Portuguese and Brazilians and in both countries literacy levels were directly related to the social, labor and economic dimensions, we base ourselves on this classification to categorize writers into social classes. Therefore, in general terms, on this scale there are three groups of occupations/social functions that define the socioeconomic stratum of individuals in Portugal, according to Magalhães (2001).2

  1. Higher stratum: nobles, ecclesiastics, professionals, clerks, notaries, teachers, big businessmen (property owners), big owners of land and real estate, capitalists, high-ranking officers, and goldsmiths were educated.

  2. Intermediate stratum: in occupations, such as small traders, farmers, tailors, masters of the trade, mester or trade officers, bailiffs, small traders, guards, low-ranking police officers and servants, the situation of literacy fluctuated, since they were commonly related to the level of elementary or secondary learning, or lack of any ability to read or write.

  3. Lower stratum: the occupations of servants, journeymen and slaves were jobs for illiterate individuals.

Social organization in Colonial Brazil, as well as formal education and exclusion of women, took place based on European standards (Algranti 1993; Veiga 2007; Vartuli 2012), and although there are some specificities regarding the possibility of social elevation provided by economic power and not only by family lineage (Faria 1998) and female literacy, due to special circumstances (Chequer 2002), eighteenth-century society was, in general, organized like this (Fonseca 2008; Julio & Vartuli 2016):

  1. Higher stratum: nobles from Portugal and their descendants, ecclesiastics, high-ranking officers, big traders, professionals, individuals who had achieved economic powers, regardless of their occupation.

  2. Intermediary stratum: free and unskilled workers; low-ranking officers; servants.

  3. Lower stratum: slaves.

3. Methodology

We adopted the theoretical and methodological contribution of Labovian Sociolinguistics (Labov 1972) for the selection, collection, and analysis of data. Furthermore, as we investigate linguistic phenomena from the past, we find support in the assumptions of Historical Sociolinguistics (Romaine 2009). We also consider the norm of abbreviations and the levels of complexity of the rules (Seixas 2020) and the social history of language in the 18th century as criteria for analyzing our data.

3.1. The variables

The dependent variable of this work corresponds to the use of the abbreviations, our research object; the standard variant corresponds to compliance with the norms of the period, as prescribed in the eighteenth and nineteenth-century normative works investigated by Seixas (2020), and the non-standard variant corresponds to any employment that is not prescribed in these works or that deviates from their prescription.

As for the conditioning factors, also called independent variables, they were divided into two groups: i) the extralinguistic (external independent variable), related to aspects external to the system (social) and (ii) the linguistic (internal independent variable), related to internal aspects of the system (structural).

First, the occurrences of the abbreviations were analyzed according to three external variables: (a) level of education; (b) socioeconomic stratum and (c) sex. Schooling levels were divided into (i) Major Studies (higher education) or (ii) Minor Studies (lower education). As for the socioeconomic stratum, its division was as follows: (i) higher socioeconomic class (A) and (ii) lower socioeconomic class (B). We included male and female informants equally in the corpus. Regarding internal variables, the abbreviations identified were also analyzed according to two variables: (a) typology and (b) complexity of the rule.

3.2. The corpus

The textual genre selected was the personal letter. This choice was due to the fact that the genre has a high possibility to be clear, simple and approaches vernacular as characteristics (Tin 2005). More specifically, in the selection, we look for thematic content of a private nature, less formal style and documents with the compositional structure of the letter. Regarding its structural composition, we consider the following classification for the identification of the textual position in which the abbreviations were found (Spina 1994; Belloto 2002): (a) the protocol (or initial protocol) – pre-textual section, which contains the dates topical (local) and chronological, addressing (to whom it is addressed) and greeting; (b) the text – it constitutes the body of the document and contains the registered fact, with the presentation and the purpose of the letter; and (c) the eschatocol (or final protocol) – post-textual section, which contains the closing and signature; may contain topical and chronological dates if they do not appear in the protocol.3

The selected missives come from two sources: (i) Barão de Camargos Collection – Historical Archive of the Museu of the Inconfidência; (ii) Post Scriptum: A Digital Archive of Ordinary Writing (Early Modern Portugal and Spain).

About the locality choices, it is important to mention that the inclusion of Brazilian and Portuguese informants in the corpus also aimed to evaluate a possible interference of the external variable locality. This variable, however, was not selected as significant by the program, what we attribute not only to an influence of European education (Veiga 2007) in the formation of literate citizens in Colonial Brazil, but also to the fact that our object of study belongs to the domain of the written language and the differences between European and Brazilian Portuguese are sensitive in the phonetics, being materialized, therefore, in the speech domain (Coutinho 1969; Câmara Jr. 1972). Therefore, we restrict ourselves to the evaluation of internal and external variables directly linked to the degree of literacy of the scribe, but we believe it is important to clarify the motivations for the constitution of the corpus.

As in the Portuguese and Brazilian societies of the 18th century there was a direct relation between social stratification and the function that the individual exercised (Magalhães 2001; Mendes 2008), we adopted the profession/occupation and the family (in the case of women who had no profession/occupation) as criteria for the characterization of writers in two social strata: (i) higher socioeconomic class (Class A) and (ii) lower socioeconomic class (Class B). Thus, we established the other equivalences when considering as representatives of class A the writers who had a profession/occupation that required training in Major Studies (higher level of education) or certain study time, such as lawyer, doctor, notary, priest, nun, among others, or, if the profession/occupation is not identified, as in the case of women, if they were from noble families or possessing assets. And, as representatives of class B, we are considering those who had a profession/occupation that did not require training in Major Studies or a determined period of study, that is, those who fit the level of Minor Studies (lower level of education), such as tailor, carpenter, cook, slave, among others, or if the profession/occupation is not identified, in which family they lived or the profession/occupation of the father/husband.

Faithful to the Labovian methodology, we selected 24 missives, which were organized as follows: (a) 3 from socioeconomic class A women, with Major Studies education or from a wealthy family, from Portuguese Language of Brasil (PLB); (b) 3 from a socioeconomic class A woman, with Major Studies education or from a wealthy family, from the European Portuguese (EP); (c) 3 from a socioeconomic class B woman, without Major Studies education or from an unwealthy family, from PLB; (d) 3 from a socioeconomic class B woman, without Major Studies education or from an unwealthy family, from the EP; (e) 3 from a socioeconomic class A man, with Major Studies education or in a specific study area, from PLB; (f) 3 from a socioeconomic class A man, with Major Studies education or in a specific study area, from EP; (g) 3 from a socioeconomic class B man, without Major Studies education, from PLB; (h) 3 from a socioeconomic class B man, without Major Studies education, from EP.4

We also determined the number of words as an element of quantitative similarity in the sample for comparing extra and intralinguistic factors. For that, we verified the letter that had fewer words (113) and, from that, the same number was considered in the others. We emphasize that, as the textual genre used has a formulaic structure and that in the protocol and eschatocol there is a considerable number of abbreviations, and as it was considered the total number of words in the letter that served as the basis for determining this value, we consider in the other letters, first, the number of words contained in these segments and then the count was made at the beginning of the body of the text, until the total of 113 words per letter was reached.

Then, we start analyzing the data. As one of the objectives was to ascertain the dis-obedience to the norm, the corpus abbreviations were identified and classified into standard and non-standard variants, according to the prescriptions found/inferred.

Subsequently, we classify and quantify the abbreviations as to the typology, according to the classification proposed by Costa (2007): 1. As a general sign, subdivided into: 1.1. Abbreviation by suspension/apocope (an. = anno; Fr. = Frei; pag. = pagina); 1.2. Initial, subdivided into: 1.2.1. Simple Initial (D. = Dom/Dona; F. = Fiel); 1.2.2. Reduplicated Initial (D.D. = Desembargadores; R.R. = Reverendos; P.P. = Padres); 1.2.3. Composite acronym or Acronym (SIDA = Síndrome da Imunodeficiência Adquirida; OTAN = Organização do Tratado do Atlântico Norte); 1.3. Abbreviation by contraction/syncope (Roiz = Rodriguez; Frz = Fernandez; Snr = Senhor); 1.4. Abbreviation by superscript letter (Illmo = Illustrissimo; Fevro = Fevereiro; p.a = para); 1.5. Mixed abbreviation (V.Ex.a = VossaExcelencia; S.Paulo = São Paulo; S. Mage = Sua Magestade); 2. Special sign (Lx.a = Lisboa; § = parágrafo; = mil); 3. Tironian or shorthand notes (.S. = Scilicet; & = et); 4. Numerical abbreviation ( = primeiro; 7bro = Setembro; 8.as = oitavas).5

The next step was to classify and quantify the occurrences according to the level of complexity of the rule to measure the employment of the scribes in relation to the more and less complex rules. This aspect was analyzed since there are some types that are simple and do not require specialized knowledge; in turn, there are others that are more sophisticated and that require specialized knowledge, whether in the discursive domain or grammatical nature. Therefore, we take as a basis the degree of legibility of the abbreviations proposed by Duchowny, Coelho & Coelho (2014), whose hierarchy involves the following continuum, in the respective order of specific knowledge required (the more to the right, the more complex): Numeric → Superscript letter → Contraction → Suspension → Simple initial → Reduplicated initial → Special sign.

The next step was the data coding for the use of the Goldvarb X statistical analysis program (Sankoff, Tagliamonte & Smith 2005).6 Subsequently, through the controlled analysis of the data, we verified which external and internal factors, measured in terms of percentages and of factor weights (FW), favor the application of the rule (in this case, the application of the standard norm, which was the reference variable) and, consequently, we judge the in-adequacy of our hypotheses.

3.3. On the sociolinguistic characterization of scribes

The degree of education of the scribes was inferred from their profession/social function and, consequently, their socioeconomic stratum. Then Chart 2 shows the socio-historical-cultural information of the scribes that make up our corpus, which, along the lines of our study, constitutes its sociolinguistic record.

Chart 2

Information about the scribes.

NAME SEX COUNTRY CITY PROFESSION/SOCIAL ROLE STRATUM YEAR
Joze de Oliveira Pinto Masculine Brazil Mariana Academic A 1766
Miguel Ignacio dos Santos Masculine Brazil São Luís Lawer A 1780–1789
Francisco da Costa Pinto Masculine Brazil Paranaguá Ensign A 1778
Joaquim Teixeira Masculine Brazil São Thiago Servant B 1782
Manoel Vasconcelos Masculine Brazil Rio de Janeiro Quadrilheiro7 B 1733–1734
Theodosio Machado Masculine Brazil Bahia8 Tailor B 1755
Domingos Vandelli Masculine Portugal Lisboa Professor A 1795
Pedro Affonso Pires Masculine Portugal São Romão Ecclesiastic A 1761
Joze da Costa Martins Masculine Portugal Lisboa Apothecary A 1757
Manoel Rodrigues Masculine Portugal Lisboa Cook B 1714
Silvestre Rodrigues Masculine Portugal Lisboa Bricklayer B 1777
Manoel Pinto Pereira Masculine Portugal Lisboa Ship crew B 1753
Clara Felicia da Roza Feminine Brazil Mariana Merchant/property owner A 1783
Thomasia Izabel Gonzaga Feminine Brazil Recife Judge’s daughter A 1770–1772
Maria Catarina da Conceição Feminine Brazil Olinda Nun A 1770
Jozefa Feminine Brazil São João Servant B 1780
Thereza de Jesus Feminine Brazil Recife Slave B 1770–1772
Izabel Gomes Veiga Feminine Brazil São Gonçalo Seamstress B 1733
Hellena Josepha Feminine Portugal Lisboa Nun A 1750
Clara Rosa Feminine Portugal Viana do Castelo Captain/miner’s daughter A 1743
Margarida Ingracia Feminine Portugal Lisboa Merchant’s wife A 1757
Thomasia dos Santos Feminine Portugal Setúbal Peasant B 1788
Domingas Rodriguez Feminine Portugal Cercosa Tailor/Carpenter’s Wife B 1737
Roza Maria Feminine Portugal Viana do Castelo Fisherman’s Wife B 1786

The round in the Goldvarb X program (Sankoff, Tagliamonte & Smith 2005) generated the following results, as shown in Table 1.9

Table 1

Round data in the Goldvarb X program.

Standard Non-standard
Occurences/Total 259/467 208/467
Percentage 55.5% 44.5%
Selected variables
  1. Typology of abbreviations

  2. Education level/socioeconomic stratum10

  3. Sex

  4. Rule complexity

467 abbreviations were collected and analyzed of which 259 (55.5%) correspond to standard use and 208 (44.5%) to non-standard use. This shows us a discreet preference of the scribes for forms that respect the norm of the period. In addition, of the controlled variables, four were selected by the program, in the following order of significance: 1) typology of abbreviations; 2) level of education/socioeconomic stratum; 3) sex; and 4) complexity of the rule. As we aim to verify the use of the norm and the influence of the variables, let us proceed to the analysis according to the dis-obedience to the norm, which will be divided into two blocks: (i) extralinguistic variables and (ii) linguistic variables.

4. Analysis

4.1. Extralinguistic variables

4.1.1. Level of education/socioeconomic stratum

The preliminary round showed us that the variable level of education/socioeconomic stratum was the external factor with the greatest significance in the use of the standard variant. In Table 2 we observe the specific data regarding the dis-compliance with the standard.

Table 2

Use of standard and non-standard variants depending on the level of education/socioeconomic stratum.

Standard Non-standard Total
N % N % N % FW11
Major Studies/Class A 172 66.7 86 33.3 258 55.2 .67
Minor Studies/Class B 87 41.6 122 58.4 209 44.8 .30
Total 259 208 467

Of the 467 occurrences of abbreviations, 258 (55.2%) were computed in the group with the highest level of education/socioeconomic class A and 209 (44.8%) in the group with the lowest level of education/socioeconomic class B. These data show that, in general, the abbreviations are more used by the most educated/wealthy. When looking at the dis-obedience to the norm, standard occurrences (66.7%) advantageously outperform non-standard occurrences (33.3%) in the group of more literate and wealthier scribes. In turn, in the group with less education, the opposite occurs, but with a less significant difference, with 41.6% of standard occurrences and 58.4% of non-standard occurrences. These data reveal that the uses are inversely polarized, with preference for the use of the standard variant in the more educated class/class A and the non-standard in the less educated class/class B, which, therefore, points to the confirmation of our expectation that the abbreviations translate the schooling level.

When we analyze the factor weight, it is the group with Major Studies/Class A that has the greatest significance for the application of the rule (FW = .67), that is, that favors the use of the abbreviation according to the standard. If we relate this result to the fact that greater contact with the norm and greater domain of the possibility of using the rules was directly linked to the higher level of schooling in the 18th century (Soares 1998; Marcuschi 2001; Mendes, 2008), it is possible to consider, under this perspective, that the individual who fits into this group has a greater mastery of the norm and, consequently, has a higher degree of literacy.12

4.1.2. Sex

In the wake of the characterization of the degree of literacy according to external factors, it is also possible to characterize the sex of the scribe. As in the 18th century women had less access to formal education (Algranti 1993; Adão 1997; Magalhães 2001), the expectation is that the male sex shows greater fidelity to the norm. Let’s observe, then, what the data in Table 3 reveal.

Table 3

Use of standard and non-standard variants according to gender.

Standard Non-standard Total
N % N % N % FW
Masculine 147 64.5 81 35.5 228 48.8 .64
Feminine 112 46.9 127 53.1 239 51.2 .36
Total 259 208 467

The first observation is that the percentage of abbreviations computed in males (48.8%) is slightly lower than in females (51.2%). These data do not allow us to ensure that brachygraphic resources are more frequent in each sex. When refining, however, the analysis for dis-compliance with the norm, we realize that men have a significant preference for the use of the standard variant, since they use it in 64.5% of the occurrences and women in 46.9%. As for non-standard employment, we observed an inversely similar profile (35.5% vs. 53.1%, respectively). The factor weight confirms the trend presented in the percentage data, since the program selected the male gender as a favorer of the use of the standard norm (FW = .64). This result is supported by the social scenario of the 18th century society: due to its social role and restricted access to educational practices, with some exceptions, the female gender was the socially excluded group and did not have a high level of schooling/literacy, even in the wealthier stratum.

These data therefore confirm our expectation regarding greater male fidelity to the norm and also signal the adequacy of the hypotheses that external factors interfere with the use of abbreviations – in this case, indicate the sex of the writer – and, consequently, that brachygraphic resources mirror the level of literacy – here, the highest level of education linked to the male sex.

4.2. Linguistic variables

4.2.1. Typology of abbreviations

As with external factors, internal factors can also give us clues about the level of education of who is behind the quill. Table 4 shows the data obtained with the correlation fidelity to the standard and typology of the abbreviation.

Table 4

Use of standard and non-standard variants depending on the type of abbreviation.

Standard Non-standard Total
N % N % N % FW
Superscript letter 99 41.9 137 58.1 236 50.5 .40
Simple initial 84 81.6 19 18.4 103 22.1 .65
Mixed 43 48.3 46 51.7 89 19.1 .48
Special Sign 27 96.4 1 3.6 28 6.0 .86
Contraction 3 42.9 4 57.1 7 1.5 .30
Suspension 2 66.7 1 33.3 3 0.6 .18
Numeric 1 100.0 0 0.0 1 0.2 *NA
Total 259 55.5 208 44.5 467

Of the 467 abbreviations analyzed, we identified seven types: 1) superscript letter, with 50.5% (236 occurrences), a very high percentage compared to the other types; 2) simple initial, with 22.1% (103 occurrences); 3) mixed, with 19.1% (89 occurrences); 4) special signal, with 6.0% (28 occurrences); 5) contraction, with 1.5% (7 occurrences), a very low percentage compared to the previous ones; 6) suspension, with 0.6% (3 occurrences) and 7) numerical, in the last position, with only 0.2% (1 occurrence). In turn, the reduplicated initial, compound acronym and tironian note typologies were not used.

Let us now deal with the factor weight: there are two significant types, the special sign (FW = .86) and the simple initial (FW = .65). Among them, it is the special sign that favors the use of the standard norm. If it is a fact that greater contact with the norm and greater mastery of the possibility of using the rules were fundamentally associated with a higher level of education in the 18th century (Soares 1998; Marcuschi 2001; Mendes 2008), it is coherent to consider the special sign as the more complex type and the simple initial as the simplest type and link the higher productivity in the use of the special sign to a higher level of education of the scribe.

It is possible to ascertain such a conjecture by contrasting, in Table 5 as follows, the use of more and less complex types in the two socioeconomic strata, with the expectation of obtaining greater productivity in the more educated/affluent group.

Table 5

Use of the special sign and simple initial depending on the level of education/socioeconomic stratum.

Special sign Simple initial Total
N % N % N %
Major Studies/Class A 32 66.7 55 49.6 87 54.7
Minor Studies/Class B 16 33.3 56 50.4 72 45.3
Total 48 111 159

As expected, the more educated/wealthy group uses, in general, more complex abbreviations than their counterpoint (54.7% and 45.3%, respectively). As it is relevant the use of the special sign as it reflects more specialized knowledge (Duchowny, Coelho & Coelho 2014) and, consequently, a higher level of education, again the more educated/wealthy group surpasses the opposing group (66.7% and 33.3%, respectively); this time, with an even more significant percentage differential of 98%.

Regarding the less complex type of use, the simple initial, we have a completely different profile than that seen previously, since the percentage difference between the less educated/wealthy group (50.4%) compared to its counterpart (49.6 %) is almost null, of only 0.8%. Therefore, it is not possible to associate a profile that favors the use of simple initials.

As with the level of schooling/socioeconomic stratum factor, it is possible to associate the sex of the scribe with the choice of certain types and justify such preference according to their level of education, given the situation of discrepancy in access to education (Adão 1997; Algranti 1993; Magalhães 2001). Taking into account, again, the distinction of knowledge dispensed in the employment of each typology (Duchowny, Coelho & Coelho 2014), and the male privilege regarding access to educational practices, it is expected that the results reflect the context of the period.

The previous speculation seems to be confirmed, as shown in Table 6 above, since the special sign, the most complex type (Duchowny, Coelho & Coelho 2014), is widely used by men (64.6%) in relation to women (35.4%) – there is an 82.4% higher usage. In turn, regarding the simple initial, the usage profile is the opposite: of a total of 111 occurrences, 57.7% were female and 42.3% male, a less discrepant percentage differential, of 36.4%. Therefore, the greater male dominance over the most complex variety is notable as well as the female dominance over the least complex.13

Table 6

Use of the special sign and simple initial according to the sex.

Special sign Simple initial Total
N % N % N %
Masculine 31 64.6 47 42.3 78 49.1
Feminine 17 35.4 64 57.7 81 50.9
Total 48 111 159

Once again basing on the social history, the preference for certain types and rules is linked to the level of education of the scribe (Magalhães 2001). Indeed, if greater loyalty to the norm and greater frequency of complex types is evident, this is an indication that the hand that carries the feather is more literate and is male. This allows us to assume, then, the use of the special sign as another criterion for measuring the degree of literacy.

4.2.2. Complexity of the rule

In Table 7 we deal with the data obtained about the interference of the internal variable rule complexity.

Table 7

Use of standard and non-standard variants depending on the complexity of the rule.

Standard Non-standard Total
N % N % N % FW
Level 1 222 47.5 204 43.7 426 91.2 .47
Level 2 37 7.9 4 0.9 41 8.8 .79
Total 259 208 467

If we consider productivity according to the level of complexity of employment in the abbreviation usage standard, those that include general rules, level 1 (91.2% – 426 occurrences), are considerably more accessible compared to those that require knowledge of rules level 2 (8.8% – 41 occurrences). This huge divergence is not surprising, given the 18th century social context and the extremely dissonant access to education (Magalhães 2001; Vartulli 2012).

Let us turn our attention to the data referring to the level 2 rule, the more complex type, because, despite its lower productivity, it is in it, as indicated by the factor weight (.79), that are the most relevant information capable of helping the researcher to trace the social profile of the scribe, especially related to the greater or lesser degree of schooling. Thus, as the complexity of the rule was proposed by Seixas (2020) as a tool to refine the identification of the rule domain, as well as the expectation, as exposed in the section on social history, that an individual with a higher level of education in the 18th century has greater domain of the norm, we will correlate, in Table 8 below, this variable to the level of education/socioeconomic stratum of the scribe.

Table 8

Complexity of the rule depending on the level of education/socioeconomic stratum.

Major studies/Class A Minor Studies/Class B Total
N % N % N %
Level 1 228 48.8 198 42.4 426 91.2
Level 2 30 6.4 11 2.4 41 8.8
Total 258 209 467

Confirming our expectation, the preference for level 1 rules is expressive in both groups (48.8% in the most educated/wealthy vs. 42.4% in the least educated/wealthy) compared to level 2 (6.4% and 2.4%, respectively). Therefore, this is another indication that the rules that demanded more formal and/or specific knowledge were much less used and, consecutively, less accessible – a portrait compatible with the generalized education inaccessibility in the 18th century (Magalhães 2001).

As we seek graphic evidence that reflects distinction in the level of education, our focus is also on the abbreviations that fall under level 2 due to its greater complexity. Despite the low percentage of use, only 8.8% of the total, we obtained relevant information: there is a vertiginous percentage differential, of 166%, since 6.4% of the occurrences came from the hands of the most educated/affluent scribes and only 2.4% from the less privileged group.

We have therefore another evidence that the most socioeconomically privileged class and, consequently, more educated in the period, employs more complex brachygraphic resources – which, in the same way, confirms our expectation based on the relation between a higher level of education, greater contact with the norm, greater knowledge of the possibilities of using the rules (Soares 1998; Marcuschi 2001; Mendes, 2008) and, consequently, a greater degree of literacy.

As previously tested, let us correlate the variable in question with the sex of the scribe.

At first, we observed in Table 9 that the most complex forms (level 2) are more recurrent in the group of men (5.9% vs. 2.7%); its counterpart, level 1, occurs almost similarly, with a slight female preference (48.4% vs. 42.8%).

Table 9

Complexity of the rule according to sex.

Masculine Feminine Total
N % N % N %
Level 1 200 42.8 226 48,4 426 91.2
Level 2 28 5.9 13 2.7 41 8.8
Total 228 239 467

As we are looking for clues that reflect literacy traits, our attention is again turned to data on level 2. Although the percentage of use is relatively low (8.8%), it is in the contrast between the sexes that we obtain revealing data: 5.9% occurred in the male group and 2.7% in the female group – that is percentage differential of 118.5%. Once again, the preference profile for the most complex variety (level 2) is linked to the group of men, as expected, precisely because of their privileged access to education.

Given the above, it is clear that the internal variable complexity of the rule has a significant role in the use of abbreviations and that those that fall under level 2, therefore, provide us with evidence of a more literate hand. And, for this reason, the abbreviations are consolidated as graphic clues that reflect the segregated and patriarchal society existing in the 18th century.

4.3. Degree of literacy: the clues

This conjunction of results showed us that the abbreviations in fact reflect social characteristics of the scribe and that, therefore, they can serve as an auxiliary resource for sociolinguistic characterization, when one is not in possession of such information. Our analysis allowed us to establish, according to the factor weight of each controlled variable, a methodological criterion that aims to help the researcher to infer, with more accuracy, the degree of literacy of the owner of the quill from the use s/he makes of the brachygraphic resources. Considering that in the 18th century there was an overlap between education level, socioeconomic factors, and sex, it is possible, taking as a parameter the scribe’s obedience to the abbreviations’ employment rules, to establish a continuum that allows the scribe to be characterized according to greater or lesser degree of literacy, as shown in Figure 1 below. The farther to the right of the continuum, the lower the degree of literacy and, consequently, the lower the social class, in addition to the high probability of being a female scribe.

Figure 1
Figure 1

Literacy continuum based on the social clues.

Source: Seixas (2020: 193).

Aiming at greater assertiveness, it is recommended to complement this continuum with the combination of formal elements, using, for this purpose, an analysis of both the type of abbreviation used and the level of complexity of the rule. Thus, the greater the use of special sign type abbreviations and level 2 complexity abbreviations, the greater the degree of literacy of the quill owner.

Despite the results being promising for the establishment of an auxiliary methodology, it should be noted, however, that the proposed categorizations indicate trends and that, inevitably, there are profile deviations, as predicted by the Labovian Theory (Labov 1972).

5. Final considerations

As we verified that each social group has a profile regarding the knowledge of the norm and that the factor weights extracted from the data reflect the significance of the variables in the identification of obedience to the norm, the abbreviations proved to be an effective auxiliary resource for the identification of social characteristics of the 18th century scribe, mainly of his degree of literacy. In addition, we found that the greater the use of certain linguistic elements, here called graphic clues, the greater the level of education of the person behind the quill.

More specifically, considering the analysis undertaken here, it is possible to state that

  1. the extralinguistic variables level of education, socioeconomic stratum and sex interfered with the use of abbreviations. Among them, we found that it is the level of education/socioeconomic stratum of the scribe that has the greatest influence on the use of the standard. Thus, when a manuscript is more consistent with the use of the abbreviation usage norm, it appears that its scribe is an educated individual and belonging to a higher socioeconomic stratum;

  2. the linguistic variables typology of abbreviation (special sign) and complexity of the rule (level 2) reflect the greatest domain of norms of the writer. Among them, we verified that it is the variable typology of the abbreviation that most favors the use of the standard norm.

In view of the significance of the results achieved and the perspective they offer for the establishment of a research methodology with corpora of past synchronies, we can safely say that this investigation has brought an important contribution to the studies that look at diachrony, especially of sociolinguistic orientation, since it solidifies the abbreviations as a linguistic object. In addition, it allows the expansion of the scope of past samples that can be adopted as corpora due to the greater possibility of recovering the scribe’s data, so esteemed by the Labovian variationist current.

The abbreviations in fact proved to be a relevant source of understanding of the phenomena of the Portuguese Language and the socio-historical-cultural aspects of the writers in the 18th century. But more than that, these graphic clues certainly hold more linguistic secrets waiting to be unraveled.

Notes

  1. We also opted to investigate works from the 19th century for three reasons: the data also represent the writing of the 2nd half of the 18th century, therefore, there is a possibility that the norm of that century is close to that of the previous century; some works took many years to complete their editing and publication process during the periods in question and could still represent the norm of the previous century; some works are editions subsequent to the first publication, which may contain the 18th century norm. [^]
  2. Magalhães (2001) points out that this is a macrodivision, which leads us to understand that all groups had profile deviations. [^]
  3. Due to the more informal character of the selected genre, we are sometimes faced with variations in the compositional construction of the letters. This is because information normally located in the initial protocol was not always written in the position intended for it, sometimes occurring in the body of the text or in the eschatocol, or else it was simply not written. [^]
  4. Four letters from Brazilian senders are from Barão de Camargos Collection and twenty letters are from the Post Scriptum archive, twelve from Portugal and eight from Brazil. [^]
  5. We emphasize that we do not consider, however, the nomenclature given by the author for the abbreviations of the word que (found in our corpus in the forms Q-, Q̃, q, q., q’ or ) as a Tyrian note – we consider as a simple initial, according to Seixas (2020) and Berwanger & Leal (2008). [^]
  6. Specific statistical program for the analysis of linguistic variation, which operates with the notion of variable rules; shows the percentage data and the factor weight to identify the in-significance of the external and internal variables on the analyzed phenomenon. [^]
  7. Humble Official of Justice. Ordered in Chamber to serve three years; takes the oath, can quote; has public faith, & carries a stick. Obliged to keep an eye on his crew, & to know if disorders are involved in it, to warn him; goes to the noises with weapons, to the pulls out, & fights, with spear, or stick; arrests the guilty, that are given in a list, & in the lairs of the powerful can seek, & arrest the men who chases (Bluteau 1728: 7–8). Original citation: Official humilde de Justiça. He ordenado em Camera para servir tres annos; faz juramento, póde citar; faz fé, & traz vara. He obrigado a vigiar a sua quadrilha, & saber se nella se commettem desordens, para o avisar; sahe aos ruidos com armas; acode aos arrancamentos, & brigas, com lança ou vara; prende os culpados, que lhe forem dados em rol, & nos coutos dos poderosos póde buscar, & prender os homiziados que seguir (Bluteau 1728: 7–8). [^]
  8. It does not mention the city, only the state. [^]
  9. The original round also had locality as an extralinguistic linguistic variable and the number of syllables of the abbreviated word and the word class of the abbreviation as linguistic variables; however, they were not analyzed in this work. [^]
  10. The level of education and the socioeconomic stratum were analyzed as a single variable because they are directly intertwined factors, since the level of education was a relevant criterion for identifying the socioeconomic class of the writer in the 18th century, apart from the exceptions. Thus, we avoid the repetition of the data, since the results we obtain in relation to the writers of each schooling level (Major Studies and Minor Studies) will be the same of the two socioeconomic strata (Class A and Class B, respectively). [^]
  11. Factor weight values above .50 are significant for the application of the rule (Sankoff, Tagliamonte & Smith 2005). [^]
  12. Bearing in mind that the correlation between schooling level and social strata in the 18th century was not a rule, since it was not uncommon for individuals from the upper bourgeoisie or intercontinental traders to have primary education or even to be illiterate, the hypothesis regarding this correlation can be confirmed with testing in other corpora. [^]
  13. We did not run the program to check the use of the typology according to sex and other variables, since the focus of the research was the use of the norm, so we do not have here the factor weight, only percentages in relation to the use of the special sign and simple initial according to sex. This, therefore, sets precedents for future studies. [^]

Acknowledgements

To CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) for the granted scholarship, which made possible the exclusive dedication to this research.

Competing Interests

The authors have no competing interests to declare.

References

Adão, A. (1997). Estado Absoluto e Ensino das Primeiras Letras: As Escolas Régias (1772–1794) [Absolute State and Literature Teaching: The Royal Schools (1772–1794)]. Lisboa: Fundação Calouste Gulbenkian/Serviço de Educação.

Algranti, L. M. (1993). Honradas e devotas: mulheres da Colônia [Honored and devoted: Women from the Colony]. São Paulo: José Olympio.

Barbosa, A. G. (1999). Para uma história do português colonial: Aspectos lingüísticos em cartas de comércio [For a history of colonial Portuguese: Linguistic aspects in trade letters]. Unpublished Doctoral thesis, Universidade Federal do Rio de Janeiro.

Barbosa, A. G. (2013). Perfis sócio-históricos de redatores do século XIX: Graus de letramento e ortografia [Socio-historical profiles of 19th century writers: Degrees of literacy and spelling]. Seminário Nacional do Projeto para a História do Português Brasileiro. Maceió, October 13–14.

Barbosa, A. G., & Lima, A. X. (2019). O controle indireto de perfis sócio-históricos em corpora histórico-diacrônicos: A identificação de graus de letramento pela grafia etimológica do século XIX [The indirect control of socio-historical profiles in historical-diachronic corpora: The identification of degrees of literacy by the etymological spelling of the 19th century]. In T. Ataliba Castilho (Ed.), História do Português Brasileiro 2: Corpus diacrônico do português brasileiro (pp. 168–205). São Paulo: Contexto.

Belloto, H. L. (2002). Como fazer análise diplomática e análise tipológica de documento de arquivo [How to do diplomatic and typological analysis of archival documents]. Projeto Como Fazer, 8. São Paulo: Arquivo do Estado, Imprensa Oficial do Estado.

Bluteau, R. Vocabulario portuguez & latino: Aulico, anatomico, architectonico … [Portuguese & Latin vocabulary: Aulic, anatomic, architectonic …]. Coimbra: Collegio das Artes da Companhia de Jesus, 1712–1728. 8v. [Online access at http://dicionarios.bbm.usp.br/pt-br/dicionario/edicao/1. Last accessed: Nov. 1, 2016].

Câmara Jr., J. M. (1972). Línguas europeias de ultramar: O português do Brasil [Overseas european languages: Brazilian Portuguese]. In J. M. Câmara Jr. (Org.), Dispersos de J. Mattoso Câmara Jr. Seleção e introdução por Carlos Eduardo Falcão Uchôa (pp. 71–87). Rio de Janeiro: Fundação Getúlio Vargas, Serv. de publicações.

Carneiro, Z., Oliveira, K., & Lobo, T. (2013). Metodologia para a identificação de perfis socioculturais de redatores em corpora históricos: Pesquisa direta em arquivos [Methodology for the identification of socio-cultural profiles of editors in historical corpora: Direct research in archives]. Seminário Nacional do Projeto para a História do Português Brasileiro, Maceió, October 13–14.

Chaves, E. (2006). Implementação do pronome Você: A contribuição das pistas gráficas [Implementation of the pronoun You: The contribution of graphic clues]. Unpublished Master’s dissertation, Universidade Federal de Minas Gerais.

Chequer, R. M. P. (2002). Negócios de família, gerência de viúvas: Senhoras administradoras de bens e pessoas (Minas Gerais 1750–1800) [Family business, widow management: Ladies that managed assets and people (Minas Gerais 1750–1800)]. Unpublished Master’s dissertation, Universidade Federal de Minas Gerais.

Coelho, S., Ramos, J., & Duchowny, A. (2015). Processos e mudanças em abreviaturas mineiras setecentistas: Regularidade e ruptura [Processes and changes in 18th century abbreviations from Minas Gerais: Regularity and rupture]. Filologia e Linguística Portuguesa, 17(2), 333–352. DOI:  http://doi.org/10.11606/issn.2176-9419.v17i2p333-352

Cohen, M. A. M. (2010). Reexame de um caso clássico à luz de novos dados: A gramaticalização e a reanálise de mente [Review of a classic case in the light of new data: Grammaticalization and re-analysis of mente]. In L. Vitral & S. M. Coelho (Eds.), Estudos de processos de gramaticalização em português: Metodologias e aplicações (pp. 57–74). Campinas: Mercado de Letras.

Costa, R. (2007). Edição semidiplomática de Memória Histórica da Capitania de São Paulo, Códice E11571 do Arquivo do Estado de São Paulo [Semidiplomatic edition of Historical Memory of the Captaincy of São Paulo, Codex E11571 of the Archive of the State of São Paulo]. Unpublished Master’s dissertation, Universidade de São Paulo.

Coutinho, I. L. (1969). Gramática histórica [Historical grammar]. 6. ed. rev. Rio de Janeiro: Livraria Acadêmica.

Duchowny, A. T., Coelho, S. M., & Coelho, G. H. (2014). Sistema de abreviaturas de documentos adamantinos setecentistas [18th century adamantine document abbreviation system]. Revista Letras, 1(90), 233–252. DOI:  http://doi.org/10.5380/rel.v90i2.36430

Faria, S. C. (1998). A colônia em movimento: Fortuna e família no cotidiano colonial [The colony in motion: Fortune and family in colonial daily life]. Rio de Janeiro: Nova Fronteira.

Fonseca, T. N. (2008). Letras, ofícios e bons costumes: Civilidade, ordem e sociabilidades na América portuguesa [Letters, crafts and good customs: Civility, order and sociability in Portuguese America]. Belo Horizonte: Autêntica.

Julio, K. L., & Vartuli, S. M. A. R. (2016). Em sua companhia: Estratégias femininas no processo educativo dos filhos, Vila Rica, Minas Gerais (1770–1830) [In your company: Female strategies in the educational process of children, Vila Rica, Minas Gerais (1770–1830)]. Educação em revista, 32(3), 91–117. DOI:  http://doi.org/10.1590/0102-4698153826

Labov, W. (1972). Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.

Magalhães, J. P. (2001). Alquimias da escrita: Alfabetização, história, desenvolvimento no mundo ocidental do antigo regime [Alchemy of writing: Initial reading instruction, history, development in the western world of the old regime]. Bragança Paulista: Editora da Universidade São Francisco.

Marcuschi, L. A. (2001). A oralidade e letramento [Orality and literacy]. In L. Marcuschi (Ed.), Da fala para a escrita: Atividade de retextualização (pp. 15–43). São Paulo: Cortez.

Marquilhas, R. (1996). Leitura e escrita em Portugal do Século XVII [Reading and writing in 17th century Portugal]. Unpublished Doctoral thesis, Universidade de Lisboa.

Mendes, S. T. P. (2008). Combinações lexicais restritas em manuscritos setecentistas de dupla concepção discursiva: Leitura e escrita [Restricted lexical combinations in eighteenth-century manuscripts with double discursive conception: Reading and writing]. Unpublished Doctoral thesis, Universidade Federal de Minas Gerais.

Romaine, S. (2009). Socio-historical linguistics: Its status and methodology. Cambridge: Cambridge University Press.

Sankoff, D., Tagliamonte, S. A., & Smith, E. (2005). Goldvarb X: A variable rule application for Macintosh and Windows. Toronto: Department of Linguistics, University of Toronto.

Santiago, H. S. (2012). Um estudo do português popular brasileiro em cartas pessoais de “mãos cândidas” do sertão baiano [A study of Brazilian popular Portuguese in personal letters of “candid hands” from the Bahian hinterland]. Unpublished Master’s dissertation, Universidade Estadual de Feira de Santana.

Santiago, H. S., & Carneiro, Z. N. (2019). Tratamento metodológico das mãos inábeis em corpora diacrônicos [Methodological treatment of unskilled hands in diachronic corpora]. In A. T. Castilho (Ed.), História do Português Brasileiro 2: Corpus diacrônico do português brasileiro (pp. 92–119). São Paulo: Contexto.

Seixas, V. C. (2018). Gramaticalização de senhor: A abreviatura como pista gráfica [Grammaticalization of senhor: The abbreviation as a graphic clue]. In S. M. Coelho (Ed.), Gramaticalização e mudança linguística (pp. 101–116). Belo Horizonte: Viva Voz.

Seixas, V. C. (2020). As abreviaturas na escrita setecentista: Pistas gráficas como recurso subsidiário de caracterização sociolinguística do escrevente [Abbreviations in eighteenth-century writing: Graphic clues as a subsidiary resource for the sociolinguistic characterization of the writer]. Unpublished Doctoral thesis, Universidade Federal de Minas Gerais.

Silva, D. C. (2004). O processo de escolarização no Termo de Mariana (1772–1835) [The schooling process in the Term of Mariana (1772–1835)]. Unpublished Master’s dissertation, Universidade Federal de Minas Gerais.

Soares, M. (2003). Alfabetização e letramento [Initial reading instruction and literacy]. São Paulo: Contexto.

Soares, M. (1998). Letramento: Como definir, como avaliar, como medir [Literacy: how to define, how to evaluate, how to measure]. In M. Soares (Ed.), Letramento: Um tema em três gêneros (pp. 68–125). Belo Horizonte: Autêntica.

Spina, S. (1994). Introdução à Edótica: Crítica textual [Introduction to Edotics: Textual criticism]. 2nd ed. São Paulo: Ars Poética/Edusp.

Tfouni, L. V. (2001). A escrita – remédio ou veneno? [Writing – medicine or poison?]. In M. A. Azevedo & M. L. Marques (Eds.), Alfabetização Hoje (pp. 51–69). São Paulo: Cortez.

Tin, E. (2005). A arte de escrever cartas: Anónimo de Bolonha, Erasmo de Rotterdam, Justo Lípsio [The art of letter writing: Anonymous from Bologna, Erasmus from Rotterdam, Justo Lípsio]. Campinas: Editora da Unicamp.

Vartuli, S. M. A. R. (2012). Senhoras da Colônia: Cultura escrita e inserção feminina em Minas Gerais (1784–1822) [Colony ladies: Written culture and female insertion in Minas Gerais (1784–1822)]. Cadernos de História. Belo Horizonte, 13(18), 32–63. DOI:  http://doi.org/10.5752/P.2237-8871.2012v13n18p32

Veiga, C. G. (2007). História da Educação [History of Education]. São Paulo: Editora Ática.

Primary Handwritten Source

• Historical Archives of the Museum of the Inconfidência (annex III – Casa Setecentista do Pilar Museum)/IBRAM/Ministry of Tourism – Barão de Camargos Collection: documentation of the Barão de Camargos family (1730 to 1898) – private correspondences, notes and receipts.

Primary Electronic Source

• Centro de Linguística da Universidade de Lisboa – CLUL (Ed.). (2014). P.S. Post Scriptum: A Digital Archive of Ordinary Writing (Early Modern Portugal and Spain). [Online access at http://ps.clul.ul.pt/. Last accessed: March 10, 2020].