“I never get a thing that ain’t been used”: A diachronic corpus-based study of second-hand consumption

1 Introduction

Although ecolinguistics has been defined and understood in many different ways, there is widespread agreement that at a very general level it refers to “the (inter-)relationships between language and (its) environment” (Couto 2014: 122). There is less consensus when it comes to setting out the specific goals of ecolinguistics. What we notice, however, is a tendency to focus on negative aspects of language and environment. In LeVasseur (2015), a summary of a survey among scholars involved in ecolinguistics, one respondent presents ecolinguistics as aiming “to interrogate and potentially expose the anthropocentric structures and usages of languages – structures and usages which may be contributing to ideology fundamentally detrimental to ecological well-being” (LeVasseur 2015: 24, emphasis added). Stibbe (2020: 2) points out that “[a]t the core of ecolinguistics is the idea that some of the stories industrial civilisation is based on are not working, because society is becoming increasingly unequal and increasingly destructive of the environment” (emphasis added). And even Fill (2018), who recognizes both the positive and negative contributions of language, seems to associate ecolinguistics more with the latter than with the former: “ecolinguistics deals with the impact of language and discourse in describing, but also aggravating and perhaps alleviating, environmental problems” (Fill 2018: 3, emphasis added).

As against this “focus on negative critique” (Stibbe 2018: 165), Stibbe shows how ecolinguistics can – and should – consider positive discourses, that is, “discourses which can inspire people to find well-being in ways that do not require over-consumption and treat the natural world with respect and care” (Stibbe 2018: 176). He gives the examples of indigenous oral cultures that promote living in harmony with nature and the Slow Food movement, which highlights the link between food and preservation of the environment. The present article seeks to adopt this positive perspective by taking as a starting point a concept that promotes anti-consumerism and eco-consciousness, namely second-hand consumption. While not a recent phenomenon, second-hand consumption is seen as a way of reducing the ecological footprint, and hence pollution, in these times of environmental crises. In an attempt to examine how discourse on second-hand consumption has evolved over the last two centuries, a diachronic analysis of the term second-hand is carried out, with particular emphasis on the connotations that have come to be associated with it.

Various methods have been applied within ecolinguistics. Chen’s (2016) meta-analysis of 76 journal articles on ecolinguistics published between 1991 and 2015 reveals that half of them are empirical studies, and that among these, a large majority apply qualitative methods based on ethnography (17 studies), macro policy analysis (13 studies), or discourse analysis (6 studies). Only 2 studies use quantitative surveys and experiments. While underlining the value of qualitative research in ecolinguistics, Chen (2016: 114) argues that “a promising direction for future ecolinguistics research is quantitative-oriented strategies involving corpus-driven or experimental methods”. In fact, the synergies between ecolinguistics and corpus linguistics have recently started to be exploited, as illustrated by Alexander (2018) or Poole (2022). The present study relies on data from the Corpus of Historical American English (COHA) (Davies 2010) and takes advantage of the quantification that is allowed by corpus techniques. At the same time, however, it observes data in their contexts of use, hence giving access to more qualitative findings as well.

The article is structured as follows. Section 2 provides some background to second-hand consumption, while Section 3 describes the corpus and methodology used. The results of the quantitative and qualitative analyses are presented in Section 4. Section 5 discusses how these results can be used to improve the representations of second-hand consumption. The conclusion is found in Section 6.

2 Second-hand consumption

Second-hand consumption has been shown to be “gaining incredible momentum” (Turunen et al. 2018). The market has risen steadily over the last few years, and it is expected to double in the next five years (ThredUP 2021: 4). The COVID-19 pandemic seems to have helped develop habits of buying second-hand among consumers, with 33 million consumers having bought second-hand clothes for the first time in 2020 (ThredUP 2021: 7). Some groups of consumers are more likely to buy second-hand, in particular the younger generations and mothers with young children (ThredUP 2021: 22–23).

There are several motives that may drive people to buy second-hand goods. One of the major motives is financial: people who cannot afford new products are “excluded from normal consumption practices” (Williams and Windebank 2002: 502) and have to turn to alternative modes of consumption including second-hand. This type of ‘forced second-hand consumption’ probably explains the “shame and stigma associated with second-hand consumption” (Franklin 2011: 156).

Yet, surveys among consumers of second-hand goods reveal many other motives for adopting this mode of consumption, such as the discovery element of second-hand shopping, the possibility of buying objects that cannot be found in mainstream shops or the history and authenticity of second-hand items (e.g. Steffen 2017). The fashionability of a certain “vintage style” (Dubin and Berman 2000) has also encouraged second-hand shopping among some consumers.

Recently, ecological motives have been gaining ground: second-hand consumption is seen as a way of reducing consumerism and hence limiting environmental impact. It has been estimated that buying second-hand instead of new reduces the carbon footprint by 82% (ThredUP 2021: 9). Second-hand consumption contributes to the circular economy, since objects that would otherwise have been thrown away are bought instead of new items and reused. In the clothing industry, for example, it is a means of curtailing the recent trend of fast fashion. More generally, it can be described as “a non-excessive, modest consumption type” (Steffen 2017: 191).

As a result of these more positive motives, the stigma associated with second-hand goods is thought to have gradually decreased among consumers, giving way to “a more confident and exuberant championing of second-hand shopping as an ethical alternative to unregulated and uncontrolled ‘consumerism’” (Franklin 2011: 156). One of the objectives of the present article is to examine the possible change in the representations of second-hand consumption by means of a linguistic analysis of the contexts of use of the word second-hand in various texts produced in the US over a period of 200 years. The frequency of occurrence of the word across the decades will also be investigated, as well as the types of second-hand items that are mentioned in different time periods.

This type of linguistic approach complements studies of second-hand consumption based on questionnaires or interviews among consumers (e.g. Guiot and Roux 2010; Steffen 2017) in that (i) it examines the representations of second-hand consumption in discourse, rather than directly asking consumers about their opinions, and (ii) it makes it possible to go back in time by using older texts as a source of information, rather than having to rely on answers provided by contemporary consumers. The quantification allowed for by the study of large quantities of authentic language data will also be put to good use, together with the more qualitative examination of the data in context.

3 Corpus and methodology

This linguistic analysis of second-hand consumption relies on COHA (Davies 2010), a collection of texts in American English produced over a period of two centuries, from 1820 to 2019, for a total of almost 500 million words. Close to half of the data in this diachronic corpus represent fiction, while the remaining texts come from magazines, non-fiction/academic books, newspapers (from the 1860s onwards), and TV/movie scripts (from the 1930s onwards). The last mentioned category is the only one that is made up of spoken(-like) data. The fact that the corpus is skewed towards fiction means that fictional representations of second-hand consumption are likely to predominate, and that these may be guided by other principles (e.g. literary creation) than genres which seek to provide information about the real world.

All the occurrences of the word second-hand, spelled with or without a hyphen and in one or two words (second-hand, secondhand, second hand),^[1] were extracted from COHA by means of Mark Davies’s web interface (https://www.english-corpora.org). 1,987 instances were retrieved, which were all examined one by one to determine whether the word second-hand was used to refer to objects that one acquires after they have been owned by someone else. Other uses of the word, such as Examples (1) and (2), were excluded from the analysis.

In total, 1,207 instances of second-hand were kept after the stage of manual disambiguation and form the basis of this study. It should be noted that, despite the large size of the corpus used, this number is relatively small, especially when considering frequencies per decade. This should be recognized as a possible limitation to the generalizability of the findings.

Different types of analyses were carried out on the data. Because of the exclusion of certain instances of second-hand, the analyses had to be performed outside Mark Davies’s interface. All the relevant occurrences of second-hand from COHA were grouped into a file that was analyzed by means of WordSmith Tools (Scott 2008). The first step was to compute the relative frequency of the word second-hand for each decade. This made it possible to examine the diachronic evolution of its frequency (see Section 4.1).

A collocational analysis was then performed, focusing on the first position to the right of second-hand. The aim of this analysis was to identify the words that are recurrent in that position, where items sold or bought second-hand are more likely to appear, as shown in Example (3) with the collocate books. This provided insights into the types of second-hand items most often mentioned in COHA, and how this evolved across the last two centuries (see Section 4.2).

Thirdly, information was sought about people’s representations of second-hand consumption in COHA. All the collocates of second-hand within a window of five words to the left and five words to the right were automatically extracted. These collocates were then compared with Hu and Liu’s (2004) opinion lexicon, a list of almost 7,000 positive and negative words meant for sentiment analysis. All the collocates of second-hand in COHA found in the list were assigned a positive or negative score, adjusted according to their relative frequency. In Example (4), for instance, among the five words to the left and five words to the right of second-hand, two words were included in Hu and Liu’s (2004) opinion lexicon, namely pride and beautiful, both of them positive and hence attributed a positive score. Since pride occurred only once as a collocate of second-hand in COHA, however, its positive score was lower than that of beautiful, which occurred twice as a collocate of second-hand in the corpus.

The identification (and quantification) of these collocates made it possible to determine the semantic prosody of second-hand, that is, the positive or negative connotations that come to be associated with a word through its repeated use with certain (positive or negative) collocates (cf. Partington 2004), and to see how this semantic prosody evolved over the last two centuries (see Section 4.3.1).

It is important to acknowledge the limitations of an automatic approach relying on Hu and Liu’s (2004) opinion lexicon, in the sense that the collocates as used in context may not actually have the positive or negative value attributed to them, which could have an impact on the scores provided in Section 4.3.1. The adjective cheap, for example, is classified as a negative word in Hu and Liu’s (2004) opinion lexicon. However, it often has a positive value in the context of second-hand, as appears in Example (5).

Another example is buggy, listed as a negative word in Hu and Liu’s (2004) opinion lexicon (probably because of its use as an adjective related to bug), but used as a neutral collocate of second-hand in COHA, to refer to a baby carriage.

In addition, the automatic approach does not necessarily detect all cases of positive or negative connotations, either because the collocates occur outside the window of five words to the left and five words to the right, as is the case with the word loathsome in Example (6), or because the words surrounding second-hand are not explicitly positive or negative, as in Example (7), which describes the defective condition of a motorcycle through words which are not emotionally loaded.

In order to compensate for these limitations, the automatic analysis of the semantic prosody of second-hand was combined with a more qualitative approach which involved manually examining all the occurrences of second-hand in context and determining those that clearly expressed positive or negative connotations. While admittedly partly subjective, this approach made it possible to identify some of the positive and negative features of second-hand consumption as expressed in COHA (see Section 4.3.2).

Before turning to the analyses proper, it should be pointed out that the frequency of second-hand and the positive/negative scores of its collocates will be provided for each decade in COHA, but that for the other results the corpus will be divided into four groups of five decades, so that tendencies across longer periods can be better highlighted.

4 The evolution of second-hand in COHA

4.1 Frequency

The word second-hand (and its spelling variants) used in the sense of ‘having been owned by someone else before’ occurs 1,207 times in the whole COHA, which represents an average relative frequency of 2.5 occurrences per million words (pmw). As Figure 1 reveals, however, this frequency varies over time. Despite some slight ups and downs, we notice a general increase from the 1820s (0.3 occurrences pmw) to the 1960s (4.3 occurrences pmw, the highest relative frequency across the period covered by COHA). The 1970s show a relatively sharp decrease in the use of second-hand, with a drop in frequency from 4.3 to 2.8 pmw. The frequency continues to drop until the 2000s (2.1 occurrences pmw) and it then goes up very slightly in the 2010s (2.2 occurrences pmw), although we would need to wait until new data are available in COHA to determine whether this is the beginning of an upward trend or not.

Figure 1:

Diachronic evolution of the relative frequency per million words of second-hand in COHA.

Although second-hand consumption could be expected to increase with economic recession and decrease with economic growth, the diachronic evolution of the frequency of second-hand does not seem to follow the economic situation in the US in this manner. Apart from a peak in the 1940s, which could arguably correspond to World War II, the so-called Gilded Age (1870–1900) and the post-World War II era (1946–1973), which both correspond to periods of economic expansion, display relatively high frequencies of the use of second-hand, including the highest frequency of all in the 1960s. By contrast, the 1970s recession and the subprime mortgage crisis at the end of the 2000s correspond to a comparatively lower frequency of the word. As for the rise in ecological awareness that has characterized the last decade, it does not seem to materialize in a significantly higher frequency of the word second-hand.

It should be borne in mind, however, that talking or writing about second-hand consumption cannot be equated with promoting this mode of consumption or actually consuming second hand. In fact, we will see in Section 4.3 that some occurrences of second-hand present second-hand consumption in a negative light. In addition, not all instances in the corpus refer to the time when the text was written, as appears in Example (8), taken from a text written in 2012 but describing events that took place in 1943. It may also take some time before consumption habits find their way into language representations, especially fictional ones (novels or movies, for example). Finally, as will be shown in Section 5, a decrease in the frequency of the term second-hand may be compensated for by the use of alternative terms referring to second-hand consumption. All these elements could explain the apparent divergence between economic and linguistic evolution.

(8)

A group of Tuskegee pilots is sent to Italy in 1943, where they patrol in secondhand planes. (COHA, 2012, Fiction)

4.3 Representations of second-hand consumption

4.3.1 Semantic prosody of second-hand

A major factor in whether people buy second-hand or not is their attitudes towards this type of consumption. In this section, we therefore examine the representations of second-hand consumption in COHA and how they have evolved over the last two centuries. To do so, an analysis of the semantic prosody of second-hand was carried out on the basis of its collocates (five words to the left and five words to the right). A positive or negative score was automatically assigned to some of them, based on Hu and Liu’s (2004) opinion lexicon and dependent on the frequency of the collocates.

Figure 2 shows the diachronic evolution of the overall positive and negative scores per decade, taking all the relevant collocates into account. While the actual scores vary in quite irregular ways, what we notice is that the positive and the negative scores in one and the same decade are very often similar and, in a way, cancel each other out. The differential score is thus equal or very close to 0 in the 1840s, 1850s, 1900s, 1940s, 1970s, and 2000s.

Figure 2:

Diachronic evolution of the positive and negative scores of the collocates of second-hand in COHA (window of five words to the left and five words to the right).

By contrast, the negative score clearly predominates in the 1960s (negative score = −0.55; positive score = 0.24). It will be reminded from Section 4.1 that this decade displayed the highest frequency of the word second-hand, which was interpreted as being somewhat paradoxical, given that it corresponds to a period of economic growth. It now appears that while people often talk or write about second-hand consumption in the 1960s, it seems to be more in negative terms than in positive terms. This is illustrated by Examples (33) and (34), which include the collocates loud and obsolete, scored negatively on the basis of Hu and Liu’s (2004) opinion lexicon, but also the adjectives old, ungodly, and sad, which could all be said to contribute to the negative semantic prosody of the two extracts.

(33)

Gene wrote that it was old and secondhand, painted a loud, ungodly green (COHA, 1962, Fiction)

(34)

– And all I got in return was a secondhand, obsolete molecular transporter.

– Indeed. How sad.

(COHA, 1967, TV/Movie)

Overall, across the whole period covered by COHA, positive collocates are more frequent than negative ones. This tendency is most characteristic of the 1890s and 1920s, and to a lesser extent the 1990s and 2010s. The 1890s and 1920s correspond to relatively prosperous decades: the 1890s were part of the Gilded Age (see Section 4.1) and the 1920s, also called the ‘Roaring Twenties’, were marked, until the Wall Street crash of 1929, by economic growth and easy access to mass-produced consumer goods. It is thus interesting that second-hand goods were apparently still spoken quite highly of in these decades. As for the 1990s and 2010s, we saw in Section 4.1 that the frequency of second-hand was lower in these decades than in the 1940s–1960s, but it now turns out that the semantic prosody of the term might have become more positive. In the case of the 2010s, this could help explain the paradox highlighted in Section 4.1 that the subprime mortgage crisis and the increasing ecological awareness did not really coincide with an increase in the frequency of second-hand: the word is used less often, but when it is used, it is more likely to be framed in positive terms.

Table 2 lists the collocates of second-hand that were automatically assigned a positive or negative score on the basis of Hu and Liu’s (2004) opinion lexicon. The collocates of five decades are grouped together to avoid too fragmentary a presentation. Only collocates occurring at least twice are listed in the table, but the many words occurring just once were taken into account in the analysis as they can all contribute to the positive or negative semantic prosody of second-hand. As pointed out by Stubbs (1995: 249), “[w]hat is significant is the summed frequency of semantically related items”.

Table 2:

Collocates within five words to the left and to the right of second-hand in COHA (minimum frequency of 2) assigned a positive or negative score on the basis of Hu and Liu’s (2004) opinion lexicon.

1820–1869		1870–1919		1920–1969		1970–2019
Positive	Negative	Positive	Negative	Positive	Negative	Positive	Negative
Good (3)	Cheap (2)	Like (4)	Cheap (5)	Boom (8)	Cheap (7)	Like (9)	Cheap (4)
		Better (3)	Buggy (2)	Like (6)	Worn (4)	Good (8)	Junk (3)
		Enough (3)	Dirty (2)	Enough (4)	Shortage (3)	Chic (4)	Worn (3)
		Well (3)	Poor (2)	Good (4)	Cheaply (2)	Enough (4)	Dark (2)
		Good (2)	Rough (2)	Famous (3)	Haunt (2)	Best (3)	Hard (2)
		Led (2)		Well (2)	Poor (2)	Available (2)
		Pretty (2)			Ragged (2)	Beautiful (2)
		Ready (2)				Glory (2)
		Worth (2)

The first finding emerging from Table 2 is that money-related collocates – both positive and negative according to Hu and Liu’s (2004) opinion lexicon – are found in the four time periods distinguished within COHA (cheap, cheaply, enough, poor, worth), as illustrated in Example (5) above as well as Examples (35) and (36) below.

(35)

Fred accumulated enough money to buy a second hand press with which he printed advertising bills for his employer and local merchants (COHA, 1997, Non-fiction/Academic)

(36)

When a man has entered on his career, if he enters on it with a will, he soon finds that all books and objects not essential as tools for his work creep stealthily into the dusty corner, or to the inaccessible top shelf of the bookcase, or if he is very poor, to the second-hand bookshop. (COHA, 1883, Fiction)

Among the positive words, good or its comparative and superlative forms are also found in the four time periods, underlining the quality of certain second-hand items, as in Examples (37) and (38).

(37)

She bought a flannel nightgown and two long-sleeved dresses – one gray pinstripe, one forest green – and she started keeping an eye out for a good secondhand coat. (COHA, 1993, Fiction)

(38)

There are some items in here that are better second-hand. (COHA, 1958, Fiction)

Particularly in 1970–2019, but also in earlier periods with lower frequencies, some collocates refer to the beauty of second-hand items, with words such as beautiful, chic, and glory. Examples include (39) and (40).

(39)

In the end, this vendor of beauty and secondhand chic stayed with us for some time. (COHA, 1945, Fiction)

(40)

Oh, there she was. In all her gleaming, second-hand glory. (COHA, 2012, TV/Movie)

Among the negative words, many refer to the used and possibly damaged nature of second-hand items or the scruffy appearance of second-hand shops, with collocates such as dirty and rough (in 1870–1919), ragged (in 1920–1969), or worn (in 1920–1969 and 1970–2019). Some illustrations are given in Examples (41)–(43).

(41)

Quick as a cat, active, lithe, he was over a six-foot fence in the rear of a building in a flash, and crouched, a black shape, against the back door of an unpretentious, unkempt, dirty, secondhand shop that fronted on West Broadway. (COHA, 1917, Fiction)

(42)

Pow bought for ten cents at a second-hand book-stall a ragged volume which gave the recipes. (COHA, 1931, Fiction)

(43)

The furniture was second-hand, worn and sufficiently ugly; a lounge, chairs, rugs. (COHA, 1942, Fiction)

Like, which is one of the most frequent collocates of second-hand in Table 2 from the 1870s onwards, is mostly used to introduce comparisons (see below), rather than as a positive verb expressing appreciation. As for boom, its frequency in 1920–1969 is mainly the result of its use as an onomatopoeia before the lyrics of the song “Second-Hand Rose”, as shown in Example (44). However, its occurrence in Example (45) is different and particularly interesting, as it provides a partial explanation for the high frequency of second-hand in the 1940s (see Figure 1), namely the shortage of certain goods during the war and the necessity of buying them second-hand.

(44)

The couples pressed close together again, swaying, barely moving in place – boom, boom, boom, boom – “Second-hand hats, second-hand clothes – That’s why they call me second-hand Rose […]”. (COHA, 1924, Fiction)

(45)

In 1942, Antiques had only 8,043 readers. But the wartime shortage of household furnishings caused a boom in the secondhand market, and Antiques boomed with it, now has 29,921 readers. (COHA, 1950, Magazine)

What this analysis of semantic prosody has revealed is that second-hand consumption is talked or written about in both positive and negative terms, often in similar amounts within one and the same decade. When the negative or positive semantic prosody predominates, this can lead to a reinterpretation of the frequency counts given in Section 4.1, as is the case for the 2000s, characterized by a lower frequency of second-hand than the 1960s but a slightly more positive semantic prosody, which seems to be more in line with the economic situation and ecological consciousness typical of these two periods. The analysis has also shown that the specific collocates may vary over time, but that certain semantic fields are shared across decades, including that of money (with both positive and negative terms), the positive fields of quality and beauty, and the negative fields of damage and scruffiness.

4.3.2 Qualitative analysis of the occurrences of second-hand

In addition to the analysis of the semantic prosody of second-hand based on its collocates and their positive or negative scores, all the occurrences of the word were manually examined in context to assess the positive or negative representations of second-hand consumption in COHA. The main observations are reported below. Note that given the more subjective approach, no attempt at quantification is made. Also, no systematic distinction is drawn between the different time periods, since the aim here is mainly to identify the positive and negative features of second-hand consumption as they transpire from the texts.

Example (46) illustrates the types of negative representations that can be found in COHA, with the subject here clearly disliking anything second-hand.

(46)

I’m a middle-aged fortune-teller who’s sick of cheap. I’m tired of secondhand, discount, giant economy-sized plastic plate cheap! I want a house, I want a mailbox and neighbors. I want an inside telephone. (COHA, 1979, Fiction)

Several reasons for such negative representations emerge from the corpus. A major one is related to the fact that second-hand goods are often described as somehow inferior: they are dirty in Example (47), in bad condition in Example (48), and possibly dangerous in Example (49).

(47)

She wore a dirty, secondhand muumuu that looked like a tent on her. (COHA, 2015, Fiction)

(48)

And the office had just a few pieces of beat-up, old, secondhand furniture and Pell’s diplomas on the wall. (COHA, 1950, Fiction)

(49)

Always buy a new seat; secondhand ones may have been in a crash or recalled. (COHA, 2002, Magazine)

This stigma can be so strongly associated with second-hand items that goods having these features are sometimes thought of as being second-hand, despite the lack of information about their origins, as in Example (50).

(50)

He examined it: the dingy wall that had been beige but needed repainting, the amateur oil portrait of Carter unframed over the mantel, the gas stove, the shelves of books piled on top of each other, the sad-looking armchairs he must have bought second hand, the studio couch with the pillows arranged over a torn place. (COHA, 1938, Fiction)

The very fact that items have been used by others before makes them undesirable in some people’s eyes, even if they present no sign of deterioration. Thus, in the lyrics of the song “Second-Hand Rose”, referred to several times in COHA, second-hand items (hats, clothes, pearls, and even Jake the plumber) are mentioned with no indication of any defect. Yet, Rose complains about having to make do with such items: “It’s no wonder that I feel abused, I never get a thing that ain’t been used”. This kind of feeling also means that there are contexts in which second-hand items, even of good quality, may be considered inappropriate, especially as gifts, as shown in Examples (51) and (52).

(51)

We suspect one of our gallant officers of the present day would not feel much flattered by being presented with a second-hand sword, however “elegant and costly” the gentleman “who had used it but little” might judge it to be. (COHA, 1861, Magazine)

(52)

Courtland hastened earnestly to impress her with the fact that Miss Brentwood was a refined girl of good family, and that it would be an insult to offer her second-hand clothing. (COHA, 1917, Fiction)

Second-hand shops and their owners often inherit the features, or at least the reputation, of the items they sell, being presented as old and dirty, as in Examples (53) and (54).

(53)

she stripped a seedy secondhand furniture store of its spavined chairs and tables. (COHA, 1951, Magazine)

(54)

One day, during a visit to the city of Buenos Ayres, I discovered in a mean street, in the southern part of the town, a second-hand bookshop, kept by an old snuffy spectacled German in a long shabby black coat. (COHA, 1922, Non-fiction/Academic)

Second-hand shops are also often pictured as messy, typically with large quantities of items being stacked or heaped in a disorderly manner and in every space of a confined area. One instance was provided in Example (12) above. Additional illustrations can be found in Examples (55) to (57).

(55)

Medium shot of the front of a second-hand book shop, with old pamphlets and magazines stacked in the front door. (COHA, 1925, Fiction)

(56)

No need to worry – he would still be there – but he quickened his pace, glanced hurriedly over the crowded array of second-hand books in the window of the Rebuilt Bookshop. (COHA, 1935, Fiction)

(57)

Rosa Lee Cunningham guided her 10-year-old grandson through the narrow aisles of the Oxon Hill thrift shop, past the crowded racks of secondhand pants and shirts, stopping finally at the row of children’s jackets and winter coats. (COHA, 1994, Newspaper)

All the negative connotations attached to second-hand consumption seem to be so deeply rooted in people’s minds that they lead to comparisons such as those in Examples (58)–(60), which exploit these connotations to describe unpleasant realities. It should be emphasized that such comparisons continue to be found in the last decade, as shown in Example (60), despite the fact that many second-hand shops nowadays offer a totally different experience, being spacious and neat.

(58)

He could feel the man looking at him though, with a stare cold and intent and yet not deliberately harsh. It was the same stare with which he might have examined a horse or a second hand plow, convinced beforehand that he would see flaws, convinced beforehand that he would buy. (COHA, 1932, Fiction)

(59)

The rest-room resembled a second-hand store. It was furnished with discarded patent rockers, lopsided reed chairs, a scratched pine table, a gritty straw mat, old steel engravings of milkmaids being morally amorous under willow-trees, faded chromos of roses and fish, and a kerosene stove for warming lunches. (COHA, 1920, Fiction)

(60)

But in masses of other poems he overdid the catchiness, and everything in the poem was so attention-getting there was no way to recall it: the purposeless glitter was packed tight like a second-hand furniture dealer’s storeroom full of chandeliers. (COHA, 2012, Non-fiction/Academic)

The negative connotations of second-hand can even lead to its use in insults such as second-hand garbage-can in Example (61) and second-hand corncob in Example (62).

(61)

Soft? Y’look like a second-hand garbage-can! (COHA, 1911, Fiction)

(62)

He’s a second-hand corncob now, I tell you. And nobody gives a damn. (COHA, 1939, Fiction)

Finally, second-hand goods and shops are sometimes associated with dishonesty (e.g. when goods are sold for more than what they are actually worth) or even illegal activities (e.g. when stolen goods are passed off as second-hand), as illustrated in Examples (63)–(65).

(63)

If you’re buying used furniture, avoid deceptive ads run by secondhand stores. (COHA, 1965, Non-fiction/Academic)

(64)

It was a waste of money, James. You shouldn’t have bought a second hand automobile. You were swindled again. (COHA, 1962, TV/Movie)

(65)

The man ostensibly conducted a little secondhand store; in reality he probably “shoved” more stolen goods for his clientele, which at one time or another undoubtedly embraced nearly every crook in the underworld, than any other “fence” in New York. (COHA, 1920, Fiction)

Turning now to the more positive representations of second-hand consumption in COHA, we can start with Example (66), which clearly shows that the subject, Holly Harris, has a favorable opinion of second-hand shopping.

(66)

A transplanted New Englander, Holly Harris, has been delighted to find that California is fertile ground for secondhand shopping. “So many people buy new, buy bigger, buy better, buy often, and they discard things just as easily,” says the editor of Rummaging Through Northern California, a Bay Area newspaper. (COHA, 1998, Newspaper)

The corpus data reveal several reasons for such positive representations. First, it turns out that, as suggested earlier, being cheap is often seen as an advantage of second-hand goods, especially for people who have limited means and/or objects that are normally very costly, as in Example (67).

(67)

But we agreed to spend the wretched trifle of the other money, left in the treasury after paying the last bills, for the largest Alvan Clark telescope that we could buy; and we were fortunate in obtaining cheap a second-hand one which came to the hammer when the property of the Shubael Academy was sold by the mortgagees. (COHA, 1872, Fiction)

The good quality of second-hand items is also sometimes underlined, as already appeared from the analysis in Section 4.3.1. This is the case in both Examples (68) and (69), despite the almost 100 years separating the two instances.

(68)

The horse having cost a good deal more than I expected to pay, I found that I could only afford a second-hand carriage. I bought a good, serviceable vehicle, which would hold four persons, if necessary, and there was room enough to pack all sorts of parcels and baskets. (COHA, 1870, Fiction)

(69)

Duane took him downstairs and proudly showed him the car – it was a second hand Mercury, nice and clean. (COHA, 1966, Fiction)

And when second-hand items show defects, they may still have sentimental value, as is clearly the case in Example (70).

(70)

The desk was her dearest possession. She had bought it from the Second Hand Furniture Mart with $11.98 of her own money. It had come to the Furniture Mart from a beach cottage and it still smelled of seaweed and salt water, with an occasional whiff of Djer Kiss and grunnion. Its stork-thin legs were stained and wobbly. It was everywhere lightly pocked with what appeared to be old buck-shot wounds but which were actually worm holes. She liked it that way. She would not have cared for unscarred varnish fresh from a furniture factory. “Oh, desk,” she said. She felt it was her real home. (COHA, 1953, Fiction)

Another positive feature of second-hand shopping, noticeable in COHA from the early twentieth century onwards, is that it makes it possible to find goods that would be difficult or impossible to buy new, such as out-of-print books in Example (71).

(71)

Shoppers come from afar, seeking a specific out-of-print book or the trendiest second-hand apparel. (COHA, 1999, Newspaper)

It also allows for chance discoveries, as in Example (72), which turns second-hand shopping into some sort of treasure hunt, an idea that is expressed several times in the corpus, including Examples (73) and (74).

(72)

A $10 pendant lamp and a painted still life, both secondhand finds, hang in the master bedroom. (COHA, 2011, Magazine)

(73)

Along the walls bordering the Retiro Park is the long line of booths of secondhand booksellers, where many a treasure may be unearthed. (COHA, 1931, Magazine)

(74)

Like many treasure hunters who scour the secondhand landscape, Bobbitt often waits until inspiration strikes to find that bargain or special item that gives character to a home. (COHA, 1998, Newspaper)

The fact that second-hand consumption is good for the environment is hardly ever expressed in COHA, and only in the more recent data. In Example (75), second-hand stores are described as an important component of our ecosystem. As for Example (76), it includes a covert reference to upcycling, with the transformation of second-hand goods into works of art.

(75)

Before you know it, it’s going to be collecting dust in those secondhand bookstores that you love so much. Those are an important part of our ecosystem. (COHA, 2019, TV/Movie)

(76)

They invest their time and money unconventionally – not in travel, cars, or gadgets, but in buying secondhand machinery to create first-rate custom crafts. (COHA, 2016, Magazine)

Another, rather subtle development in the last few decades is that some people seem to consider second-hand consumption as a natural option among different modes of consumption, leading to a mixture of first-hand and second-hand purchases, depending on what can be found where, as illustrated by Examples (77) and (78).

(77)

She’d fixed up the room with that identity in mind: locally purchased secondhand furniture mixed in with a scattering of folk-art pieces from the Antipodes and the northern territories (COHA, 2002, Fiction)

(78)

Every new thing they could afford side by side with every second-hand piece they could scavenge, and all connected. (COHA, 2006, Fiction)

Example (79) is quite characteristic, in this respect, since the author is disappointed and surprised that second-hand shopping was not mentioned in an article on “finding best deals for children’s clothing”.

(79)

I was disappointed and surprised to see that in the tips for finding best deals for children’s clothing, second-hand options were not explored. (COHA, 2009, Magazine)

To summarize this qualitative analysis of the representations of second-hand consumption in COHA, we have observed strong negative representations based on some purported inferiority of second-hand goods, which are presented as being dirty, possibly damaged, and sometimes of illegal origin. Second-hand shops are often depicted as crowded and messy, and their owners as old and bleak. Such representations seem to be so firmly anchored in the collective imagination that they are exploited in comparisons to describe unpleasant places or objects and even in insults with reference to people. The positive representations in COHA tend to emphasize the cheapness of second-hand goods and their good quality. Second-hand shopping is also described as a kind of treasure hunt which makes it possible to acquire objects that one would not have been able to find otherwise. The recent data include more matter-of-fact characterizations of second-hand shopping, seen as one among different modes of consumption, which suggests that it has become part of certain consumers’ habits. Ecological arguments, surprisingly, are hardly mentioned explicitly in the corpus.

5 Towards practical applications

Although this study has focused on a word that is, ecologically speaking, positively loaded, and although the corpus data have revealed instances that underline the positive features of second-hand consumption, it also appears from the analysis of COHA that the term second-hand can be associated with a certain stigma. This stigma presumably originated in the early days of second-hand consumption, when this mode of consumption was a necessity for people in need, but it seems to have survived until today, despite the fact that for some people second-hand consumption has become a conscious, ecological choice. The comparisons found in the corpus between second-hand items/shops and some unpleasant realities seem to be based on deep-seated stereotypes which are part of the collective imagination. The problem is that discourse reinforces such stereotypes by depicting second-hand shops as dark and messy places or by creating insults with the word second-hand.

The internet is full of articles demonstrating the ecological value of second-hand consumption, but as long as this mode of consumption continues to carry negative connotations in discourse, some consumers may be unwilling to take the plunge. Worryingly, mainstream shops sometimes exploit the disadvantages of second-hand shopping to promote first-hand shopping. In 2021, for example, a French online retailer specializing in the sale of cheap branded products launched an advertising campaign portraying ecologically minded consumers. In one of the ads, a woman is heard saying that buying second-hand goods is ecologically responsible, but that sometimes she cannot find anything she likes and that in such cases she can buy cheap new items from the retailer. Through this ad, the retailer acknowledges the ecological value of second-hand shopping, thus emphasizing its ecological consciousness and that of its customers. At the same time, however, it underlines one of the disadvantages of second-hand shopping, namely the reduced product range (in the sense that, for instance, each piece of clothing typically comes in one size only), in order to encourage their customers to turn to the new items sold by the retailer (which are presented as ecologically responsible too). In effect, this ad could therefore be said to promote first-hand consumption while riding the wave of ecological sustainability.^[2]

A corpus analysis like the one carried out here can bring to light the stigma associated with the concept of second-hand consumption, but it can also highlight positive connotations attached to it and be a first step towards improving its image. This corresponds to the goal of Positive Discourse Analysis (PDA) according to Stibbe (2018: 171), namely “to assemble clusters of linguistic features that can be useful in conveying new and beneficial stories”. In the case of second-hand consumption, the comparison with treasure hunting, already well represented in COHA, could for example be exploited in campaigns promoting second-hand shopping. Visual representations of spacious, clean, and tidy shops would also be crucial, given the rather negative image of second-hand shops that continues to be conveyed through language. More generally, (prospective) consumers of second-hand goods should not be given the impression that they are “excluded consumers” (Williams and Windebank 2002: 501) and that second-hand consumption is a “deviation from the norm” (Waight 2013: 201). On the contrary, second-hand consumption should be offered as “the new normal” (ThredUP 2021: 55) and consumers should be encouraged to “think secondhand first” (ThredUP 2021: 68). A few instances in COHA already point in this direction, as we saw earlier.

Another way in which second-hand consumption could have its image enhanced, if it turns out that the term second-hand is too negatively loaded and that this cannot be changed, would be to favor other terms with a more positive image. In fact, such terms already exist.^[3] One particularly successful alternative is the word vintage, whose use is normally restricted to relatively old items (clothes from the 1960s, for example) and whose very definition implies high quality. As appears in Figure 3, its frequency in COHA has increased steadily over the last three decades and has become much more frequent than second-hand (which could explain the unexpectedly low frequency of the latter over the last few decades). An analysis of its collocates on the basis of Hu and Liu’s (2004) opinion lexicon also reveals predominantly positive scores and an increase in positively loaded collocates over the last three decades, with words such as fabulous, gorgeous, super, savvy, remarkably, or recommend, as illustrated in Example (80).

Figure 3:

Diachronic evolution of the relative frequency per million words of vintage in COHA.

The word thrift, also used to refer to second-hand consumption, is a very positively loaded term etymologically speaking, since it is related to an Old Norse word meaning ‘wealth and success’. It currently has a frequency close to that of second-hand according to COHA, although it was more frequent in the 1980s. However, the automatic analysis of its semantic prosody with Hu and Liu’s (2004) opinion lexicon reveals an overall negative score from the 1980s to the 2000s (with collocates such as discouraging or marginally) and a very slightly positive score in the 2010s. Interestingly, the American hip hop duo Macklemore & Ryan Lewis released a song called “Thrift Shop” in 2012, which became a big hit. This song promotes thrift shopping, with lyrics such as “One man’s trash, that’s another man’s come-up” or “I wear your grandad’s clothes, I look incredible”. It also exploits the comparison with treasure hunting (“I’m hunting, looking for a come-up, this is […] awesome”) and underlines the uniqueness of second-hand items by criticizing mainstream consumption (“having the same one as six other people in this club is a hella don’t”). The message conveyed by this song is quite different from that of the song “Second-Hand Rose” dating from the 1920s which is referred to several times in COHA. Incidentally, the fact that some second-hand shops reported an increase in sales after the release of the song “Thrift Shop”^[4] suggests that popular culture may have a role to play in the promotion of second-hand consumption.

Finally, new terms could be created altogether, or borrowed from other countries. In Australia and New Zealand, for example, the term op-shop, a short form of opportunity shop, is used to refer to second-hand shops. Its focus on the chance discovery element of second-hand shopping is likely to evoke more positive images than a word like second-hand.

6 Conclusion

This study has relied on a large diachronic corpus of American English to investigate the evolution of the representations of second-hand consumption in discourse from a quantitative and qualitative point of view. Through its focus on an ecologically positive word, second-hand, it has sought to make a positive contribution to the field of ecolinguistics. Although the corpus analysis reveals that the word has decreased in frequency since the 1960s and that it carries some negative connotations which continue to be noticeable in the more recent data, there are still reasons to be optimistic about the results.

The first reason to be optimistic is that the data also include instances where second-hand carries positive connotations. The automatized analysis of its semantic prosody has shown that positive collocates actually outweigh negative ones in the 1990s and 2010s. The qualitative analysis has highlighted several positive features of second-hand shopping that make it an attractive mode of consumption for some people. The very slight increase in frequency in the 2010s could also be the beginning of an upward trend.

Next, while a detailed analysis of other terms related to second-hand consumption is beyond the scope of this article, it appears that a word like vintage has become more widespread and has been increasingly associated with positive connotations over the last few decades. It also has an intrinsically positive value, since its definition often includes a reference to high quality. Other terms such as thrift shop or opportunity shop could help improve the image of second-hand consumption too.

It should also be underlined that corpora give access to representations through language, but that the way people act and think may be different. Fictional works, which constitute almost half of COHA, could for example describe second-hand shops as dark and messy to create a certain kind of atmosphere. This does not necessarily mean that the authors of these works disapprove of second-hand shops. Reports such as ThredUP (2021) show that second-hand sales have increased over the last few years and predict that this trend will accelerate in the near future. On the internet, there are a plethora of articles extolling the virtues of second-hand consumption, and more and more celebrities are proud to say that they shop second-hand.

Finally, the kind of corpus-based analysis carried out here can be the starting point for real-world action. Once we are aware of the stigma that continues to be attached to second-hand consumption in certain types of discourse, we can create new discourses (and even new terms) that emphasize its positive features and conjure up more pleasant images. Campaigns could be organized that promote second-hand shopping by exploiting some of the positive collocates or comparisons discovered in corpora. If consumers keep hearing and reading that “second-hand clothes are cool and thrift shopping is nice”,^[5] they are more likely to try second-hand shopping for themselves and to do so with a positive attitude. The more they keep hearing and reading about second-hand shopping, the more likely they also are to consider it a normal mode of consumption. And if second-hand shopping were to lead to overconsumption, as some people start to fear,^[6] one could adapt messages to the consumer accordingly. Words have power and if they are used wisely, they can help change the world.